Understand Termodinamics

DOCUMENT RESUME
ED 344 744
AUTHOR
TITLE
SPONS AGENCY
PUB DATE
CONTRACT
NOTE
PUB TYPE
EDRS PRICE
DESCRIPTORS
IDENTIFIERS
SE 052 442
Lewis, Eileen Lob

The Deve:'.opment of Understanding in Elementary
Thermodynamics.
National Science Foundation, Washington, D.C.
Apr 91
MDR-8470514; MDR-8850552
62p.; Paper presented at the AnnJal Meeting of the
American Educational Research Association (Chicago,
IL, April 3-7, 1991).
Reports - Research/Technical (143) -Speeches/Conference Papars (150)
MF01/PC03 Plus Postage.
*Cognitive Development; Concept Formation; Grade 8;
Interviews; *Intuition; Junlor High Schools;
*Learning Processes; Middle Schools; Misconceptions;
Physical Sciences; Pretests Posttests; Science
Education; *Secondary School Science;
*Thermodynamics
Middle School Science
ABSTRACT
This study investigates how students participating in

the same curriculum construct understanding in elementary
thermodynamics during a semester-long eighth-grade physical science
class. Two questions were a0dressed: (1) How does the learners'
understanding change during the study of elementary thermcdynamics?
and (2) What role do students' intuitive conceptions play in the
restructuring and reorganization of their knowledge? Within each,
middle school students' (n=180) knowledge restructuring was
considered on an individuEl basis with generalizations made to larger
student groups. All students responded to a two part pretest and
posttest and 33 participated in all 5 clinical interviews. Three
types of students were identified. "Converging" student were quick to
recognize that heat flow is a model for thermodynamics. They use this
powerful model and continue to add new information to their existing
knowledge, finding that new knowledge is consistent with and
reinforces the way they think about thermodynamics phenomena.
"Progressing" students combine new information at the level of local
knowledge, but do not integrate the pieces of that knowledge to build
a more robust and cohesive view of thermodynamics. These students are
progressing towards the coherent understanding of the students in the
previous category. A third group of students can best be described as
having an "oscillating" perspective of thermodynamics. These students
combine experiences sporadically, change their views without
additional evidence, and finish tne course with a group of isolated
ideas. Unlike the students with a progressing view of thermodynamics,
these students do not gain more predictive ideas as time goes on, but
simply move from one set of ideas to another. These findings
contribute to understanding both the nature of the learner and the
nature of the learning process and suggest ways to design science
curricula to facilitate robust student understanding. (Author)
U.S. DEPARTMENT OF EDUCATION

Office ot Educatronal Research and improvement
'PERMISSION TO REPRODUCE THIS

MATERIAL HAS BEEN GRANTED BY
EDUCATIONAL RESOURCES INFORMATIO4

CENTER (ERIC)
a This document has been reproduced as

received from the person or orgenization
Eileen L. Lewis
originating it
Minor changes have been made to improve
reproduction quality
"%lent do nOt necessarily repreSent official
OERI position or policy
ket
The ERIC Facility has assigned

this document for processing
to:
In our iudgment, this document

is also of interest to the Clear
inghouses noted to the right.
Indexing Should reflect their
special points of view.
Points or view Of opinions stated in this docu
'414
SCOPE OF INTEREST NOTICE
TO THE EDUCATIONAL RESOUHCES

INFORMATION CENTER (ERIG)
The Development of Understanding in Elementary Thermodynamics
ri4
Eileen Lob Lewis

Science & Mathematics Education (SESAME)
University of California, Berkeley
This paper was originally presented at the American Educational Research Association's Annual
Meeting in Chicago, IL in April, 1991, Symposium 29.51, "The Nature and Development of
Scientific Understanding."
Acknowledgements:
This material is based upon research supported by the National Science Foundation under grants
No. MDR-8470514 and MDR-8850552. Any opinions, findings, and conclusions or
recommendations expressed in this publication are those of the authors and do not necessarily
reflect the view of the National Science Foundation.
We wish to express our gratitude to Doug Kirkpatrick for his spirit and assistance in all aspects of
our CLP research. We appreciate contributions of software and equipment from Bob Tinker and
Technical Education Research Centers (TERC). We gratefully acknowledge the donation of
equipment and software from Apple Computer, Inc.
5$.1
iit
PTV
_
A
Development of Understanding
INTRODUCTION
This study investigates how students participating in the same curriculum construct
understanding in elementary thermodynamics during a semester-long eighth-grade physical science
class. In particular, it addresses two questions: (1) How does the learners' nnderstanding change
during the study of elementary thermodynamics? and (2) What role do students' intuitive
conceptions play in the restructuring and reorganization of their knowledge? Within each of the
above questions, middle school students' knowledge restnicturing was considered on an individual
basis with generalizations made to larger student groups.
Three types of students were identified. "Converging" students were quick to recognize
that heat flowl is a model for thermodynamics. They use this powerful model and continue to add
new information to their existing knowledge, finding that new knowledge is consistent with and
reinforces the way they think about thermodynamic phenomena. "Progressing" students combine
new information at the level of local knowledge, but do not integrate the pieces of that knowledge
to build a more robust and cohesive view of thermodynamics. These students are progressing
towards the coherent understanding of the students in the previous category. A third group of
students can best be described as having an "oscillating" perspective of thermodynamics. These
suidents combine experiences sporadically, change their views without additional evidence, and
finish the course with a group of isolated ideas. Unlike the students with a progressing view of
thermodynamics, these students do not gain more predictive ideas as time goes on, but simply
move from one set of ideas to another. These finding contribute to understanding both the nature
of the learner and the nature of the learning process and suggest ways to design science curricula to
facilitate robust student understanding.
tile heat flow model used in the CLP curriculum was developed and refined over successive semesters. It is similar
to the historic caloric model, but stresses that heat lacks mass. Further information can be found in Linn (1988).
Theoretical Framework
The Nature of the Learner

A fundamental premise of this research is a constructivist view of learning (Driver & Bell,
1986; Linn, 1987; Linn, Layman & Nachmias, 1987; Osborne & Wittrock, 1983; Resnick, 1983;
Wittrock, 1974). There has been widespread agreement that learning science is considered to be an
active, constructive, and cumulative process in which th.; learner plays a critical role (diSessa,
1983; Glaser, 1990; Linn, 1987; Osborne & Wittrock, 1974, 1983; Shuell, 1986, 1988 ; White
and Tisher, 1986). This constructivist perspective builds on the views of Dewey (1902), Piaget
(1972), and Schwab (1973) and describes learners as experiencing the world and constructing
understanding from those experiences. As shall be seen in this and other studies, learning is
indeed cumulative and learning outcomes depend up the learner's prior knowledge as well as the
way new information is elaborated and related to the learner's existing knowledge.
The limitations of our perceptions of physical events and the complexity of the processes
we daily observe often results in learners constructing fragmentary and not infrequently
contradictory explanations of these experiences (Chaiklin, 1984; diSessa, 1983; 1987; Lewis,
1987; Lewis & Linn. 1989, April). An example can be feund in Linn (1988) where she notes that
observing motion in a friction-filled universe led individuals to reach the same conclusions as
Aristotle rather than the accepted principle of "bodies in motion tend to remain in motion." She
further notes that since objcts usually move in the direction in which they are pushed, observers
tend to ignore the speed and direction an object may already bring to the situation; they tend to
prexlict its future motion only considering the new force acting upon it.
Learners acquire a knowledge of "school science" and often are able to apply concepts only
to situations that directly mirror their school experiences. However, when varying real-world
situations are used, students revert to their pre-instruction explanations. While teachers and
researchers may be frustrated by the existence and tenacity of the individual constnictions students
bring to instruction that are contrary to the accepted principles of science, these constructions
emerge as a result of the very methods that should be fostered in scieree studentsnamely, the
ability to make observations, synthesize the results of these observations, and generalize to other
events.
Part of this study, therefore, looks at how students integrate new and existing knowledge.
Beginning with the initial conceptions of the learner, the manner in which students revise their
knowledge is investigated. This knowledge revision occurs during a 13-week thermodynamics
curriculum in which students' understanding changes as a result of epirical evidence that they
help create and actively interpret. The focus of this study is on the process of change. This
approach is consistent with other studies examining conceptual changes in knowledge (Carey,
1985, 1988; Chi & Ceci, 1987; Ranney, 1987; Shuell, 1982; Strike & Posner, 1985).
Students' Existing Knowledge

Research has demonstrated the role students' existing knowledge plays in constnicting
generalizations of science instruction to the natural world (diSessa, 1988; Driver, 1981; Gunstone,
April, 1988; Lewis & Linn ,1989,April; Linn, 1987; McCloskey, 1983; Resnick, 1987; Songer,
1989; Tiberghien, 1980, 1982; West & Pines, 1985). Ausubel, Novak, and Hanesian (1978)
were among the first to recognize the important role prior knowledge played in the learning
process. Today there is a body of liteiature that details these individually constructed knowledge
structures, their origins, contextual dependence, cueing conditions, tenacity, and pervasiveness
(Caramazza, McCloskey & Green, 1981; diSessa, 1983, 1988; Gunstone, April, 1988; Law,
1988; McCloskey,Caramazza & Green, 1980; Posner, Strike, Hewson t Gertzog, 1982).
Many of these knowledge structures, which will be called "intuitive conceptions" in this
study, are remarkably uniform among adolescents and non-scientist adults. This suggested to
Ey lon and Linn (1988) that there are some well-defined mechanisms for their creation. In a later
paper, Linn, Songer, Lewis, & Stern (1990) suggests one such mechanism. When students
acquire knowledge as a result of their experience, this knowledge is at first represented as "action"
knowledge or simply responses to situations. They cite Baldwin (1894) and Piaget (1952) in
suggesting that these actions are constnicted in isolation ofien by imitating the actions of others.
As an example, students learn to "turn up the heat," if they are cold or take off a jacket if they are
warm.
As a part of making sense of the world, students actively amalgamate their action-oriented
knowledge into intuitive conceptions. Thus intuitive conceptions are firmly grounded in
observation and personal experience. Linn, et al. (1990) state that intuitive conceptions arc
constructed by students using a conscious process mediated by language in order to "group actions
and observations into meaningful conceptions that they can use to make predictions about the
world." Intuitive conceptions represent students' conscious identification of ideas that apply to
several actions and related observations. Thus when students say, "metals airract, ansorb, or hold
cold" they are combining their experiences touching metals in cool or cold environments into a
generalization that provides some explanation for their experience. This intuitive conception gives
students excellent ability to predict the feel of metals in a variety of cool or cold situations. While
this predictive power is limited, intuitive conceptions are useful for daily living. As a first level
Development of Understunding
amalgamation of actions and observations, they provide no I, ;1 of explanation beyond
observation.
A similar view of learners' early efforts to construct meaning is found in diSessa's (1988)
phenomenological primitives (p-prims). He describes p-prims as minimal abstractions of common
phenomena. As minimal abstractions, they can be used to explain other phenomena and they are
not themselves explained within a learner's system of knowledge. Instead they are perceived to be
self-explanatory. These ideas are all consistent with research findings on the constructive nature of
the learner. It seems quite reasonable that learners should generate their own explanations and
mechanisms for observed phenomena since the mechanisms underlying many everyday
phenomena are not readily observable. Lewis & Linn (1989, April) gives a list of students'
intuitive conceptions in elementary thermodynamics. Table 1 shows these intuitive conceptions.
Lewis and Linn additionally discuss how students construct these intuitive conceptions as a result
of everyday experiences, the cultures2 in which they exist, and the use of language within that
culture.
This resistance of intuitive conceptions to change may, in fact, be a very useful and
important part of the overall learning process. Linn (1983) suggests that "persistence is useful for
advancing knowledge and that if reasoners changed all their ideas afer each contradiction, they
would gain little insight into a problem." It seems reasonable, then, that an effective mechanism
for facilitating the development of understanding may be to strengthen those parts of a student's
knowledge that are the beginning of good science, using them as anchors for conceptual
refinements, reformulations, and ultimately coherent and robust models of scientific processes
(Clement, Brown & Zietsman, 1989, March). It would also be desirable to facilitate student
experiences that encourage the development of new intuitive conceptions consistent with scientific
principles which may be cued more readily in the future. There is evidence that intuitive
conceptions tardy disappear, but the development of new, scientifically acceptable conceptions
may bolster a student's understanding and sense of success. As a result, the new conceptions
might be cued more commonly by external events. As will be show later, intuitive conceptions
consistent with scientific principles can be used by a student to construct a principled
understanding of thermodynamic phenomena.
2Culture here refers to the practices and assumptions of daily activities in student's lives. This includes family
practices, peer activities, and advertising that fosters incorrect interpretation of phenomena.
Dia=
Intniflye_fdactatim prototypic Circumstances auprising_Faknaigni
Metals as
Insulators
Metals absorb, hold, or

attract cold in cool
environment
Touching a variety of objects,

all at room temperature or in
a cool/cold environment
Use of aluminum foil for

wrapping cold objects in
order to keep them cold
Metals absorb, hold, or

attract heat in warm/
hot environment
Touching a variety of objects,

all it the same temperature
a warm/hot environment
Use of aluminum foil for

wrapping hot objects in
order to keep them hot
A wooden spoonfmsulator
standing in boiling water
Insulator as intermediate
to best material to allow
sensation of warmth from
an object.
Conductors conduct heat

more slowly than
insulators
Insulators
8S
Insulators conduct heat

fast and heat leaves so
Conductors insulators don't feel hot

Insulators absorb/trap
heat
Hot Wool
Touching a variety of objects

all at room temperature
Wool warms things up Surround any object in wool
Wool cannot be used as an

insulator for cold objects
since it warms thin s u
Table 1. A list of intuitive -nceptions, their characteristics, prototypic circumstances,

and surprising extensions fin which the concept might have evolved and to which it
applies
Process of Knowledge Integration

The persistence of intuitive conceptions in spite of a variety of instructional methods is well
documented (Clough & Driver, 1985; Erickson, 1980; Gunstone, April, 1988; Lewis, 1987;
Posner, et al., 1982; Ranney, 1987; Tiberghien, 1980, 1982). Several studies can give insight
into reasons for the persistence of intuitive conceptions. What students consider possible and
reasonable overrides tendencies to look at evidence and patterns of evidence (Tschirgi, 1980; Linn
& Pulos, 1983; Schaub le, 1988). These include the observations made of events, the
interpretations offered for such observations, and the strategies students use to acquire new
knowledge. Exposure to coherent, even meaningful knowledge may be insufficient if that material
is incompatible with existing intuitive concertions. The result may be rejection of the new
information or the creation of new, unintegrated pieces of knowledge.
Schauble (1988a, 1988')) noted that the driving force behind the students' experiments was
outcome rather than determining causal relationships. As a result, she found support for Piaget's
(1969) observation that smdents typically fail to construct experiments in which one variable at a
time is changed. The lack of controlled experiments greatly limits discovery of causality.
Conversely, successful learners in Schaub le's study consciously constructed and coordinated the
relations between evidence and their evolving theories.
Experiments and simulations provide the opportunity to experience new situations and
expand knowledge. Equally important, they provide a means for testing the ideas students have
often acquired as a result of limited experiences which permitted inconsistencies to go unnoticed.
One way to increase the likelihood of learners effectively using experimentation is to frame
experiments so that single variables are considered, intuitive conceptions are engaged, and
mechanisms for facilitating useful notions of causality are present. As will be seen, the Computer
As Lab Partner (CLP) Curriculum used in this study promotes experiments and simulations that
limit mploration to a single variable in any given experiment or simulation. More information
about the CLP curriculum can be found in the methods section of this paper and in Lim & Songer
(in press) and Linn, et al. (in press). Additionally, the simulations f real world situations were
designed to engage students' intuitive conceptions, to allow them to test their ideas, and to facilitate
the construction of scientific principles from their results (Lewis, Stern, & Linn, 1990).
Student Reflection. The role of student reflection in all of these processes must be
emphasized. Many students are not accustomed to the level of reflection required for meaningful
learning and will, in fact, resist "thinking about things for that long." Embedded in all these
considerations must be mechanisms for fostering this reflection. One such mechanism is to require
that students make written and graphical predictions before all experiments and simulations.
Further reflection occurs when students are required to resolve those predictions with their
experimental/simulated outcomes. This process for enhancing student reflection was evolved and
successfully used in the Computer As Lab Partner (CLP) Curriculum described by Linn & Songer
(in press).
Summary. In summary, ihe premises of this research are that learners construct their own
systems of knowledge and that learning environments can facilitate the construction of robrst and
coherent knowledge. In order to encourage the construction of such knowledge, the learning
environment must provide time and support for active construction. It must allow learners to
participate in activities that provide opportunities for discovering the causal nature of the world.
Those activities must engage their intuitive conceptions and provide coherent, appealing
alternatives that emphasize differentiation and integration. While the focus of all these activities
must be conceptual, the contexts must be varied and include real world phenomena in order for
learning to be robust and meaningful.
This position is supported by research describing a number of important conditions that
facilitate the development of understanding. Posner, et. al. (1982) and Strike and Posner (1985)
state that students must be dissatisfied with their current conceptions in order to alter them.
Additionally, any new concept must be both understandable and plausible. Finally, the new idea:,
must be seen as beneficial. The curriculum and experimental treatments proposed here meet each
of these requirements and provide additional opportunities for students to extend their experience to
real world processes. This provides additional support for the usefulness of their evolving
concepts while decontextualithig their application.
Method
Subjects. The subjects in this study were 180 eighth grade students (ages 12-14 year:, old) in a
middle class, ethnically diverse middle school. Each was a student in a semester-long physical
science class in which thirteen weeks were spent studying elementary thermodynamics tr
; the
CLP curriculum. The classroom teacher is a "Mentor Teacher" who has been recognized for his
excellence in teaching. A detailed description of the classroom environment can be found in Linn
and Songer (1988). All students (N = 180) responded to a two part pretest and posttest and 33
participated in all five clinical interviews.
Curriculum. The CLP curriculum used during this study is a 13 week, microcomputer based
study of thermodynamic properties and variables (Kirkpatrick, 1987; Linn, Layman, & Nachmias,
1987; Linn, & Songer, 1988; Mokros, & Tinker, 1987). Using real-time data collection and
simulations, "research groups" of students behave as working scientists in an inquiry-based
learning environment. They plan experiments, predict outcomes, use computers to display
empirical or simulated results, analyze their data, and discuss surprising or contradictory
outcomes. A detailed study of the evolution of the CLP curriculum can be found in Linn and
Songer (in press) and Linn, et al (1990). The curriculum was designed to encourage the
development of integrated understanding of elementary thermodynamic process and the variables
that affect those processes. Computer simulations of real world phenomena have been added to
engage stndents' intuitive conceptions and encourage integration of their real-time data experiments
with real world experiences. Eight Macintosh computers were used to construct the electronic
laboratory notebook and perform simulated experiments described in Lewis, Stem & Linn (1990).
Design
Since the goal of this research was to chart the development of students' understanding in
elementary thermodynamics, the research design needed to assess both group and individual
change. As a result, pretests, short tests, and posttests were administered to the entire student
population. From this sample, 36 students were selected from the middle 50% of students using a
straL e.ed random design for individual case studies. There were six students from each of the
physical science classes in the case studies. The strata used in this study were sex and class
period. This was done to obtain equal numbers of males and females and equal representation
from each period in the physical science classes. The limiteci number of males falling into the 50%
criterion in two of the class periods overly constrained choices and resulted in unequal distribution
of females and males in half the classes. This unequal class distribution wis necessary to allow for
equal nun* -as of males and females in each treatment group.
Class
ikiiigs to determine the middle 50% were based on partitioning the first pretest into
two parts by ty.,xt of question (e.g. essay, fill in, multiple choice) and variable considered. The
middle 50% from this analysis was then matched with the middle 50% from the second pretest.
This was done to ensure internal validity as well as validity between tests. The numerical scores
represent the number of total number of correct responsts on each pretest No statistical diFerence
were found between any groups as shown in Table 2.
Development of Understamiing
Table 2: Measurement of Between-Test Validity flr Two Pretests: Mean, Standard Deviations,
and Range for Interview Group, Middle 50% and Whole Population
Group
Pretest, Part 1 Mean

(Std Dev.)
Ran le
Middle 50% of Students
27.8 (2.7)
27.7 (2.6)
Whole Population (N=180
26.2 (5 6)
23-32
22-32
11-39
Interviewed Students
Pretest, Part 2 Mean (Std

Dev.)
Ran . e
34.0 (2.7)
32.0 (5.7)
29-39
20-43
15-43
31.4
5.4)
The top 25% of students were omitted from the case studies since they are often able to
integrate and generalize knowledge without assistance. The omission of the lower 25% from case
studies was made because students' placement in that percentile may be related to language skills
(in this particular setting many students are learning English as a second language), motivation, or
behavior problems. Additionally, any knowledge reorganization or reformulations promised to be

more interesting for this average group of students. If knowledge can be gained on the design of
instructional systems that are effective for the average students, they should also meet success with
students of greater abilities.
Data Sources. Since the investigation of complex processes like the development of
understanding demands a variety of data sources, this study used two different kinds of sources
for its data: (1) pretest, short tests, and posttest, and (2) 5 clinical interviews with each of 36 case
study students. Over the course of the 13-week curriculum 3 students were lost from the 36 case
study students. Two moved from the area and one student did not want to participate, all too late
for them to be replaced. Additionally, one student did not take the posttest, so that while the
number of students interviewed was 33, posttest data were only available for 32 students.
Instruments
Written Tests. A two-part pretest was devised to assess students' conceptions in the area of
thermodynamics. The questions used were refined over severai semesters ir the CLP curriculum
and have been shown to provide reliable infoimation on students' undry4nding (Linn & Songer,
ve-test reliability to be
in press). The two parts of the test were given ten days apart to alio'
established (see figure 1 for sample questions from pretests) for gru 4 selection. Several
questions on each part of the test focused on the students' understanding of heat energy,
temperature, thermal equilibtium, and variables that affect heating and cooling processes such as
starting quantity, starting temperature, insulation and conduction, and surface area.
11
10
Figure 1: Sample Questions from Fall 1989 Semester Pretest/Posttest'

la) You want to keep a soda cold for your school lunch. What is the best things to wrap it in?
b) What is the main yeason for your answer?
2a) In general, are heat energy and temperature the same or different?
(circle one)
same
b) What it: the main reason for their similarity or clifference?
different
Give an example that explains your answer.

3a) Give an example of a good conductor.
b) What makes the substance you picked a good conductor?
4a) Give an example of a good insulator.

b) What makes the substance you picked a good insulator?
5a) A metal spoon and a wooden spoon were put into a 65C oven for 2 hours. What do you predict their
temperatures will be after two hours in the oven?
temperature of the metal spoon
temperature of the wooden spoon
b) What is the main reason for your answer.

6. If a metal plate and a Styrofoam plate are in the same room, what will happen?
a) (check one)
the metal plate will be warmer.
the Styrofoam plate will be warmer.
both plates will be equally warm.
b) Give the main reason for your answer.

7 . You arc shoveling snow without wearing gloves. (Picture showing both shovels)
a) Which shovel handle would you rather use? (circle one)

metal-handled shovel
wooden-handled shovel
either shovel handle would he okay
b) Fill in the blanks to make a principle that applies to these shovels.

If two objects that differ only in material are in the same serrounding, heat energy will flow into and out of the
object. that is the better
faster than ii will flow into and out of he object
insulator / conductor
that is the poorer
insulatot / conductor
8. You take a bag of frozen peas from the freezer and place it in a ck-qtainer of hot water. (Picture showing cluster of
cold peas and container of hot water then peas in container of water.)
a) What can you say about the heat energy of the cold peas after it is placed in the hot water?
(check one)
Heat energy from the cold peas will flow into the hot water.
Heat energy from the hot water will flow into the cold peas.
Cold energy will flow from the cold peas into the hot water.
Cannot predict.
b) What is the main reason for your answer?

c) If you left and came back several hours later, what would you find?
(check one)
the peas would be colder than the water.
the peas would be warmer than the water.

the peas and the water would be the same .:mperature.
cannot predict.
1 Print and size have been significantly reduced and pictures are missing due to space limitations.
11
Questions w.re chosen to provide diversity, include both school and real world problems,
elicit students' intuitive conceptions, and display inconsistenees in student's knowledge. Many
questions were open-ended, allowing students to express their beliefs in their own language.
Interrater coding =liability for both the Pretest and Posttest was 98%.
A two-part posttest was given. The posttests were identical to the pretests except that the
fffst posttest contained 5 additional questions on heat flow. This allowed comparisons with pretest knowledge and the nature of student explanations. The results of the two pretests and posttest
were combined to produced single pretest and posttest scores.
Clinical interviews. A series of 5 clinical interviews were conducted with each of the 33
students in the case studies (see Figure 2 for sample questions). The interviews were modeled
along the lines of Piaget's "clinical method" (1969, first published in 1929). In keeping with
Piaget's method, questions were asked informally and repeated as necessaiy. In addition, the
experimenter often paraphrased responses so that a confirmation or denial of the experimenter's
perceptions of subject's response could be made. Open-ended follow-up questions were posed in
the language of the subject.
The initial interview focused on clarifying students' responses on the pretest in areas of
interest (thermal equilibrium, heat flow, and insulation/ conduction). The final interview focused
on the students' responses on the post-test in areas of interest (thermal equilibrium, heat flow,
insulation/conduction, heat energy and temperature distinction). Since a goal of this study was to
monitoring the learners' development of understanding and factors that affect any restructuring or
reorganization that occurs, interviews were conducted at four week intervals throughout the
students' 13 week study of thermodynamics. The questions were designed to assess robust
knowledge. As a result, questions were framed in familiar contexts to cue intuitive conceptions.
The questions were also designed to illustrate conflicts in a subjects' conceptions as well as to
demonstrate their ideas of causality in thermodynamics.
As a means of studying the effect of various factors on the development of understanding,

students were asked what they were thinking about, which of their ideas they thought had changed
and why. All questions required prediction and explanation to elicit the diversity of student beliefs
and conceptions and to measure the robustness of their responses. Most questions were set in the
real world to test students' ability to generalize their ideas The same questions were posed to all 33
case study students.
12
Figure 2: SAMPLE INTERViEW QUESTIONS

1. You and your friend are sharing a milk shake. Because you aren't very hungry, you and your
friend divide it up as you see.
What can you say about the heat energy present in each of your
glasses of milk shake? Does one glass of milk shake have more
heat energy than the other or are they the same? Why do you
think that? What is heat energy? Do cold things have hrg.t
energy? Were the milk shakes the same temperature when they
are poured into the glasses?
Friend Yours
2. Write the names of each of the following materials on the line where you think they belong.
Assume all the materials have the same thickness.
good
conductor
good
insulator
metal
wool
wood
Styrofoam
ceramic floor tile
glass
paper
saran wrap
Why did you place (each material) here on ue continuum line? What makes this material a good
conductorfmsulator? What does it do as an insulator? Does it work for hot/cold things only? How
did you come to understand about this material? (Probe student's understanding of conduction and
insulation.)
3. In a chemistry lab students were drying equipment in an oven like this one. The temperature of
the oven was 150C. In the oven were metal spatulas, glass beakers, and asbestos pads that had
been there overnight. What do you predict the temperature of each is? Why? (Probe student's
understanding of thermal equilibrium. What is the source of their understanding.) If you could
touch them, would they feel the same? Why? (Probe student's understanding of conduction and
insulation using the students' terms.)
13
Data Analysis
This investigation relied on two different sources of data: (1) pretest and posttest from the
entire population (N = 180), and (2) clinical interviews for case studies (N =33). The pretest and
posttest allowed statistical distribution, parametric, and nonparametric analysis of students' ability
to apply thermodynamic concepts and thereby measured the impact of instruction over the 13-week
curriculum.
Clinical Interviews. The clinical interviews were evaluated using a method developed by
Erickson (1979, 1980). Using this method transcripts are carefully examined for expressions that
state or infer the underlying beliefs of the subject. An inventory of ideas held by each subject is be
generated and forms the basic unit of analysis. These ideas are defined as follows: "an attempt by
the subject to explain a phenomenon or in some way account for a judgment or prediction the
subject makes in the course of the interview or thought questions." Attention is also paid to the
cueing conditions for these ideas, their systematically, and where their knowledge breaks down
(Posner, et al., 1982).
Level of Explanations Analysis. The Level of Explanations Analysis was developed to

measure how students' inventory of ideas changed over time. It considered the quality of an
explanation from the level of intuitive conceptions to target conceptions using the progression of
students' ideas developed by Lewis (1991). The assessment included students' use of examples as
well as their ability to interpret complex and ambiguous real world situations. Additionally, the
level of explanations scale, shown in Fig= 3, measured the consistency and cohesiveness of
students' explanations within a given interview. Since each interview was assessed individually
and different aspects of the same concept were often considered from interview to interview, there
was some fluctuation in student responses. Student responses were charted in the areas of
Insulation/ Conduction, Thermal Equilibrium, Heat and Temperature Distinction, and Heat Flow
over the sequence of tests and interviews. Figure 4 provides a diagrammatic representation of the
target conceptions in this study.
14
Figure 3: Level of apPanations Used to Analyze

Students' Interview Responses
Level of ENplanationa
1Intuitive Conceptionsresponse to questions and explanations consist primarily of
intuitive conceptions, e.g. choice of aluminum foil to wrap a cold soda because foil
holds the cold in; wool wrapped around a cold object would cause it to warm up faster
than an unwrapped object.
2Encoding New Facts without Explanationsstudents' responses display the

encoding of new factual data without an understanding of that information or the ability
to apply it, e.g. after a class experiment, students become convinced that objects in the
same environment are the same temperature. However, they cannot explain why they
are the same temperature or why they feel different to the touch.
3Mixed Predictions, Idiosyncratic Explanationsstudents at this level give some

correct and some incorrect predictions which were typically cued by the surface features
of a problem. Their explanations appear to result from an attempt to preserve intuitive
conceptions, but explain experimental outcomes, e.g. a student might state that metals
feel cold in a cold environment because they don't really absorb the cold, it just sits on
the surface of the metal. Insulators, on the other hand, were said to absorb cold into
them so that they do not feel cold to the touch.
4Mixed Predictions, Explanationsstudents make both correct and incorrect

predictions and give explanations that are a mixture of both correct and incorrect
conceptions. Incorrect explanations are usually tied to initial intuitive conceptions.
These mixed explanations exclude any idiosyncratic explanations.
5Good Predictions, Mixed Explanationsstudents make excellent predictions and give

explanations that are a mixture of target conceptions, intuitive conceptions, and
intermediate conceptions
6Target Conceptionspredictions and explanations are consistent with target

conceptions.
A plot of students' level of explanations over the sequence of tests and interviews
suggested three categories of student development. These categories are defined as "Converging,"
" Progressing," and "Oscillating." Converging Students progressed through the level of
explanations and were found to use target conceptions in their discussions of
insulation/conduction, thermal equilibrium, heat and temperature distinction, and heat flow, but
all improved their level of explanations over time. Progressing Students varied greatly in their
range of responses and did not construct target conceptions in all desired areas. They did,
however, improve their level of explanations over time. Oscillating Students oscillated between
levels of explanation or utilized idiosyncratic explanations in the areas measured. To ensure
15
reliability each Level of Explanation analysis was performed twice with a two month interval
between analyses. The interview transcripts and pre/posttest alone provided the data sources for
each analysis. Results from the use of the Level of Explanations criteria were consistent with
qualitative measures used for case studies to characterized the nature of students reasoning.
e slowly
reach
until
Insulators
uilibrium
hed
alriet)k
flows rap4dl through
flows slo
flows from higher temperature

to lower temperature
rel
differs in amount needed

to raise temperature of equal amounts
of different substances
Wool
Styrofoam
Rubber
through
rate of flow is
proportional to
temperature
differences
Temperature
Diagram of Target Concepts in Elementary Thermodynamics
Figure 4: Diagram of Target Concepts in Elementary Thermodynamics

The Level of Explanations Analysis for each of the 33 case studies allows the generation of
a within-subject summary of the growth of each individual in the sample over the course of the
semester. Subsequently, these within-subject summaries of individual growth become the basis
for the between-subject analyses. This allows inferences to be made from individual and group
development of understanding that may have broader application in instruction. A Level of
Explanations figure for each student is found after the descriptions of their category.
Case Studies. The data from pretests and each series of interviews were used to construct a case
study for each student that chronicled their development of understanding over tilt.: sequence of
tests and interviews. Attention was given to who noted any conflict in a student's responses and
how that conflict was resolved. Notation was made of what characterized the learner's responses
over the sequence of interviews. Careful attention was paid to determine ihe coherence of a
17
16
Lzielopment of Understanding
student's knowledge as well as his or her ability to apply that knowledge in a variety of settings.
The case studies were also used to create a level of explanations chart for each student.
RESULTS AND DISCUSSION

The cognitiv... growth of individuals is presented along with generalizations that can be
drawn from observations of that growth. Analysis of student interviews suggests that students fall
into three general categories of cognitive srowth: Converging, Progressing, and Oscillating.
"Converging" students are characterized by a robust and coherent understanding of heat flow as a
model for thermodynamics. They continue to add new information to their existing knowledge and
find that new knowledge is consistent with and reinforces the way they think about thermodynamic
phenomena. Their conceptions of elementary thermodynamics begin to converge with accepted
scientific conceptions. Seven out of thirty-three (or 21.2%) students fell into this "converging
category."
The second category can be described as "progressing." Students with a progressing

understanding of thermodynamics are incorporating new information to build a more robust and
cohesive view of thermodynamics based on the heat flow model. These students are progressing
towards but have not reached the coherent understanding of the students in the previous category.
Eighteen out of thirty-three (or 54.5 %) students fell into the "progressing" category.
The third group of students can best be described as having an oscillating perspective of
thermodynamics. These students oseillate between more and less predictive views of
thermodynamics and lack any ubstantial integration of their ideas. Unlike the students with a
progressing view of thermodynamics, these students simply move from one set of ideas to another
without gaining a more predictive view over time. Nine out of thirty-three (or 24.2 %) students
fell into the oscillating category.
More detailed descriptions of each of these categories of students are presented below in
addition to a case study for each category of student development. An analysis is presented
showing the interaction of posttest scores for each of these student categories.
Categories of Learners
Students' pretests, interviews, and posttest were assessed using the six Levels of
Explanations presented in the methods section of this paper. These analyses suggested three
natural categories in students' development of understanding. These categories resulted in
grouping students into "Converging students," "Progressing students," and "Oscillating students."
The results of these analyses are presemed for 6 of the interviewed students in this study, two from
1a
17
each category (see Figures 5-10) after the discussion of their category. These figures display in
tabular and graphical form the Level of Explanations used by students over the sequence of tests
and interviews for the concepts of insulation/conduction, thermal equilibrium, heat energy and
temperature differ ntiation, and heat flow. It is easy to distinguish Converging students,
Progressing students, and Oscillating students by graphically following their progress over the
sequence of tests and interviews.
Converging Students. Converging students were so named because of their robust and
coherent use of heat flow as a model for understanding elementary thermodynamics was
converging with accepted scientific views. A key characterization of the reasoning of converging
students was their ability to resolve contradictions in their responses, reason through their in and
out of class experiences and decide which explanations were most consistent with other pieces of
their knowledge. Their attention to their own inconsistency and their efforts to resolve conflicting
ideas resulted in greater reflection and some form of resolution. The following excepts provide
examples of these processes.
Student PE5 entered the curriculum with some explanations at an action level and others
representing typical intuitive conceptions. This is demonstrated in his pretest resr9nse to the
question, "You want to keep a soda cold for your school lunch. What is the best thing to wrap it
in? What is the main reason for your answer?" His reply was aluminum foil and his reason was,
"lots of people use it to keep their sodas cold." In Interview 1, he added:
PE5: "Every time you see someone with sodas, they always have it wrapped in aluminum foil, never anything
else."
E: How does aluninum foil keep them cold?
PE5: Because it's sort of like a little metal, cause metal really keeps the heat in or cold or whatever.
E: So metal works to keep things hot and cold?
PE5: I think so. I guess.
In a later pretest question he was asked to agree or disagree with the statement, "Things that
help hot objects stay hot also help keep cold objects cold," and give an example of a situation that
illustrated his answer. CE5 agreed and cited aluminum foil as an example of something that keeps
things hot and cold. During Interview 1, he agreed with his pretest response. However, on the
next interview question, he makes statements that will cause him to reconsider a later question.
E: This question asked for an example of a good conductor and you left blank.
CE5: I don't know.
E: Do you know what a conductor is?
CE5: Isn't it something that lets the heat in? All the air would come in or out or something like that.
E: So something wrapped in a conductor would stay cold if it was cold to begin with or not?
CE5: Uh, let me think...I think it would probably make it hotter, I guess. If you wrapped it around it.
18
E: If you wrapped a conductor around something cold, would it warm up faster?
CE5: Yeah, I think so.

E: Why do you think that?
CE5: If it lets in all the air, then it will make it a little hotter cause all the energy goes.
He applies this new understanding in a question at the end of Interview 1. The question
asked which dish of lasagna would stay hotter, one with a metal lid or one with a pottery lid.
E: On this lasagne question, you said the one with the metal lid would stay warmer. Can you say a little more
about that?
CE5: Oh, wait. ithink this is all wrong now. I understand this now. think this could be honer (points to the
dish with the pottery lid) because the metal, it feels hotter because all the heat went inside that, inside the lid.
While all the heat just sort of bounced off the pottery lid so it keeps the lasagna hotter. So then the lasagna
loses energy because it goes into the metal lid. That's why this (points to the metal lid) feels really hot and this
(points to the pottery lid) doesn't feel as hot.
This is a excellent example of a student combining pieces of knowledge. However, in the
next question, the typical separation of heat and coolin nrocess is demonstrated as well as the
persistence of belief in aluminum foil for keeping cold uojects cold.
E: Here you said aluminum foil was excellent for keeping hot things hot and cold things cold. Is that OK?
CE5: I think it would be between poor and good for hot and be excellent for cold.
E: So it is different for different things?
CE5: Yeah.
E: Why did you change your mind?
CE5: Well, I guess cause that's why the aluminum foil is always hot when you take it off cause it just absorbs all
the air, I gums.
E: Absorbs all the air?
CE5: Well, just like the energy of the air, the degrees or something like that.
E: Would the same thing be true if you wrapped some thing cold in aluminum foil?
CE5: Oh yeah, it would be the same. These two would both be between poor and good. Yeah, yeah. I see now.
Yeah.
He then applies his new understanding.

E: Here you said Styrofoam was good for keeping things hot and poor for keeping things cold.
CE5: Uh, I think this would be good, too. (points to his previous i-or for keeping things cold)
E: So Styrofoam is good for both keeping things hot and keeping things cold?
CE5: Yeah.
E: Better or worse than aluminum foil?
(7E5: I think it's better now.
E: If you had an aluminum cup and a Styrofoam cup and you poured the same amount of cold orange juice into
them, which orange juice would be colder 5 minutes later?
CE5: The one in the Styrofoam.
E: Why would the juice in the Styrofoam cup be colder?
CE5: Cause, let me think. Whenever you take an ordinary Styrofoam cup, it just feels like normal, not like hot or
cold. But then, because the metal absorbs all the things like I said and so you take some, it will take some cold out
of the orange juice, so the orange juice will go up a little.
While his explanation about taking some cold out needs refinement, he is exhibiting the
testing and comparing processes that are desirable in student reasoning. His understanding of
19
Development of Understa Iding
terms and use of language will improve as will his process of comparing new pieces of knowledge
to existing ones and co'rstructIng a coherent system of knowledge. By Interview 3 he illustrates an

excellent understanding of thermal equilibrium in his response to the objects in the laboratory oven
question. He stated, "150C. All of them. The metal will be quicker getting there but then this
one (points to glass) will take a little longer but nothing can get higher." He then explained how
objects could not get hotter than their surrounding. He explained why they feel differently. A later
question about a wrap for a cold candy bar illustrates his increasing understanding of heat flow.
E: You like vay cold candy bars and keep them in the freezer. You want to take one to school with you. What
would be good to wrap it in to keep it cold?
CE5: Wool.
E: Why?
CE5: It's a good insulator. It will keep all the energy inside, it won't take any energy out or bring it in.
E: Is there cold energy inside the candy bar?
CE5: No, there is heat energy but not much.
E: So what would happen?
CE5: Ie3 not letting any of the energy of the candy go out or any energy from outside go inside the candy.
E: What kind of energy would go out?
CE5: Actually energy can't leave horn it, it can't cool down. It will warm up.
E: What is the wool doing?
CE5: Preventing the warm air from coming in.
E: Does that make sense?
CE5: Yeah.
E: What if I wrapped it in aluminum foil?
CE5: It would warm up to the temperature of the room quicker.
E: Why?
CE5: The heat energy from outside can go through the foil easily.
Converging students actively used evidence from classroom experimentation/simulations and

discussions to develop new concepts and attempted to link those new ideas with their existing
ones. They were more able to subject their intuitive conceptions to examination and comparison
with their experimental/simulated outcomes construct new understandings of previous experiences.
An illustration of this is found in excerpts from interviews with Student ME7. On the pretest
question about the wooden and metal spoons in a 65C oven, her written response gave the
temperature of the metal spoon as 50C and the temperature of the wooden spoon as 65C. Her
explanation was, "The wood won't absorb the coldness, the metal spoon will." Like many
students she assumed that 65C was cool. During Interview 1 when was asked if her answer was
OK, she replied:
ME7: Yeah, I think so. I forgot about the Celsius. I thought this was Fahrenheit.
E: On the Celsius scale, this would be a warm oven.
ME7: I'd say they'd both be the same temperature, 65.
E: Why?
ME7: Bmause with the experiments in cla - that we've done. Uhm, it showed us that it doesn't matter how they
feel, it's like how they conduct the heat and stuff like that.
(skip ahead)
20
ME7: (Before the experiments) I would have said they were different temperatures.
E: But now you think they are the same?
ME7: Yes.
E: Why does metal feel cooler?
ME7: Because the mataials in the metal and the wood is different.
(skip ahttd)
E: This question about the metal and the Styrofoam plate in the same room. You said, "metal plate warmer because
assuming the room was warm and that metal heat up fast." Is that OK?
ME7: It would feel warmer. I don't know if it would be warmer. (She feels a piece of metal.) Well, this feels
colder now. Wait. It probably depends on what the room temperature was and would there be something on the
plate.
E: The plates would just be off a shelf.
ME7: You couldn't really tell, because you don't know what the room temperature would be. Cause it could be like
a really hot room and the metal plate would be warmer--it would feel warmer. But then if it was a really cold
room, then the metal plate would feel colder.
Her experience with the Probing Your Surrounaings Experiment3 enabled her to change her
understanding of thermal equilibrium. While she still doesn't understand why objects feel
different, she has incorporated her experiences into her new thermal equilibrium and generalized to
how materials feel in different surroundings. Later in the interview, OP considered the question of
whether orange juice would stay colder in an aluminum cup or a Styrofoam cup. She decided the
Styrofoam cup would work better to keep orange juice cold and hot chocolate hot. She decided to
change her pretest responses which said aluminum foil was excellent for keeping hot objects hot
and cold objects cold.
When converging students used analogies in their reasoning pmcesses, they did not rely on
surface similarities but reasoned from uncle:lying thermodynamic processes. The explanations for
predictions in the interviews ?is reflected a generalization in that students explanations were
increasingly given in general (principled) terms which the student then applyed to specific
situations. This is exemplified in the following excerpts from interviews with Student CQ6, a
foreign student who had some difficulty expressing hilisself in English. In the first excerpt from
Interview 2, he was explaining how the objects in the laboratory oven would be the same
temperature because, "They've been left in there for a long period of time, gradually, they'll get
there." He then explained that they would warm up at different rates, "Metal will be faster than
glass." He was then asked if the objects would feel the same since they were the same
temperature?
CQ6: No, the metal will feel hotter than asbestos.
E: What makes this feel hotter?
CQ6: It maybe a better conductor.
E: What is it conducing?
CQ6: ...Your hand could be conducting, but this is hotter (points to metal object) so it is conducting the heat from
that (metal object) to your hand. The asbestos is not as good of a conductor.
3This is a powerful experiment in which students predict the temperature of objects in their surroundings then
measure their temperature with a thermal probe.
21
E: Did you always think of things this way?

CQ6: Not really, doing all the experiments and the simulations. Classroom discussion are reuy helpful and people
bring different things iike real life examples...and the metal is actually the same temperature and it surprised me.
The classroom discussion helped but not as much. The principles helped. They am kind of hard.
E: Do you use the principles?
CQ6: Yeah, like, well, so I fmd out about other things that L..It kind of changes my thinking about things in
general.
He had less difficulty expressing himself in writing. This is evidenced by his response to the
heat energy and temperature distinction question. The main reason he gave for heat energy and
temperature being different was, "heat energy is the amount of heat in an object. Temperature is
just how cold or hot it is. If you acid the same amount of heat energy to different liquids, the heat
energy is still the same but the temperatures are different." His example .. as, "if you add the same
amount of heat energy to a cup of water and a pool. The heat energy contained in both of them are
the same, but when you measure the temperatures the cup of water will be higher than the
swimming pools (sic) since the cup has less mass." In a later question about the temperature of a
hot horseshoe placed in water, he stated, "If an object is placed in a surround,the object and the
surround will eventually reach the same temperature." His definitions for insulators and
conductors were equally lengthy and generalized. A last example comes from his response to a
question about a block of wood and metal on a warm "hot plate." Students were asked to draw a
picture of tell how they thought the blocks became hot. His response was, "The heat from the hot
plate comes from the generator underneath. It is conducted through the plate into the blocks. Each
will heat up at different rates. The metal will heat up faster than the wood since it's a better
conductor."
Overall characterisdcs of Converging students:
Converging students appeared to add a new piece of knowledge and integrate it with their
eidsting knowledge. This process results in a gradual reorganization and reformulations of their
exiting knowledge. A driving force in this process was a need for coherence which induced
additional reflection upon their ideas and experiences. This increased reflection and desire for
coherence facilitated knowledge reorganization and reformulation,
Converging students relied on evidence to make predictions and as a result were the first
to abandon intuitive conceptions which were contrary to scientific findings. Their explanations for
predictions were more causal in nature throughout the course of the interviews.
Converging students developed generalized explanations and increasingly expressed
explanations in principled terms. They were, however, quite able to apply those generalizations to
specific situations wiLli very different surface features. Seven of thirty-three students (or 21.2%)
22
fell into the category of converging students. Further analysis of why students fell into these
categories will be presented later in this section.
Converging students represent an ide l for the development of understanding. Figures 5

and 6 provide Level of Explanations analyses for two Converging students. A case study of a
Converging student MQ5 can be obtained from the author by request.
23
Developntsnt of Understanding
Level of Explanations Over Sequence of Tests and Interviews for

Converging Student CE5
Figure 5:
7-
Insulation/Conduction
Thermal Equilibrium
5-
Heat & Temperature

Differentiation
4-
Heat Flow
Level of Explanation
3-
Intuitive Conceptions
Encoding New Facts without
Explanations
Mixed Predictions,
Idiosyncratic Explanations
Mixed Predictions,
PreT
int #1
Int #2
Int #3
Int #4
PostT
Sequence of Tests and Interviews
Explanations
Int #5
Good Predictions, Mixed
Explanations
Target Conceptions
Level of Ex lanations:
Insulation/
equence orrests
and Interviews
Pretests
terview #
Interview it/
nterview #
Conduction
-I
4
5
11-5R-i7-4---i
6-
Interview#5
Thermal
Equilibrium
Heat 177riperature
S
5
frea757v
Distinction
Posttk.,;7-------Z6
24
Figure 6: Level of Explanations Over Sequence of Tests and Interviews for

Converging Student MQS
--er----
Thermal Equilibrium
Heat & Temperature
Differentiation
Heat Flow
1
Expianations
Mixed Predictions,
Iciosyncratic Explanations
Mixed Predictions,
Explanations
PieT
Int #1
IM #3
Int #2
Int #4
PostT
Int #5
Sequence of Tests and interviews

Explanations
Target Conceptions
Level of Ex ianations:
Thliquence o ests
and Interviews
etests
Interview #1
Interview #2
Interview #3
Interview #4
--East
#
terE-747S--v
eat : emperature
s . no
Conduction
Equilibrium
Distinction
4
4
5---4
6
' eat
ow
i
5
1
7g4
5
,
25
The following categories of students illustrate ways in which the processes displayed by
converging students breaks down.
Progressing Students. Students with a progressing understanding of thermodynamics are

incorporating new information to build a more robust and cohesive view of thermodynamics based
on a heat flow model. These students are progressing towards the coherent understanding of the
students in the previous category. The progressing category, quite naturally encompasses great
diversity in student understanding. It includes students who have just begun to make sense of their
experiences and fmd general principles underlying a variety of phenomena as well as students who
are very close to developing the level of explanations required to fit into the category of converging
students. All students in the progressing category did, however, differ in their processes and/or
time lines for developing understanding.
The reasoning processes of Progressing students is best characterized by struggles to

resolve experimental/simulated outcomes and to develop relationships between those outcomes.
They advanced by fits and starts, they often reached plateaus, followed unfruitful paths or have, in
some cases, maintained two separate bodies of knowledge (e.g. heating processes are different
from cooling processes), but overall continued to progress toward a robust and coherent
understanding of thermodynamics. This is illustrated in the following excerpts.
Student MQ8 began the CLP curriculum with common intuitive conceptions about
insulators, conducted s, aluminum foil, and thermal equilibrium. On her pretest she stated the metal
spoon would be hotter than the 65C oven and the wooden spoon would not be as hot as the oven.
During Interview 1, she agreed with her written statements about the temperature of the spoons.
Later in the interview she was asked:
E: On this question about the metal and Styrofoam plates in the same room you checked that the Styrofoam plate
was warmer and said, "The metal plate will be cooler than the Styrofoam plate because that's what I tMnk."
(MQ8 laughs) Is that OK?
MQ8: Well, actually, the experiment (refers to Probing Your Surroundings experiment), now that I think about it,
it did change the way I think because, in the experimmt, well, before I thought that, that they coald be different.
Like if they're placed in the same room, they could be different temperature, but that's not necessarily true
because they're both at the same temperature. So they both could be the same temperature but just feel different.
E: So the experiment make you think about that.
MQ8: Uh-huh.
In discussing her response to whether a metal or wooden handled shovel would be best for
shoveling snow, she was asked whether the shovels would be the same or different temperatures if
left out overnight in a snowbank.
MQ8:... Uhm, probably differun. The wooden handled one would be warmer and the metal one would be colder.
26
E: So they'd be different temperatures, and would they feel different?
MQ8: Uh huh! (laughs)
In Interview 2 she consistently applies the concept of thermal equilibrium and appears to
be developing understanding about the natuze of insulators and conductors. She predicted that all
the objects in the ski cabin would be the same temperature and explained:
Because the furniture were all the things that were in the cabin. Adjust to the room temperature because
they can't be hotter or colder because there is nothing that can make them stay... they just go to room
temperature because that is the way they are.
Do you always think that way?
E:
MQ8: I never really thought about it.
What makes you so sure that this is true?
E:
MQ8: All the experiments we did.
E:
Any in particular?
MQ8: We are always talking about the table top or the table leg, the table leg feels colder because it is metal and
metal is a better conductor to conduct heat away from your body or heat away rapidly but still, it is still room
temp., it is just whether it is a conductor or insulator. So a can be the same temperature
So did you always think that that is the way it was, or not really?
E:
MQ8: Kind of, yeah.
But it wasn't that big of a deal?
E:
MQ8: No, not really.
So you think all the objects in that cold cabin is going to be the same temperature
E:
MQ8: Basically, unless some are wrapped or in boxes... then they may be warmer.
But all these stuff that was sitting out will be the same temp.?
E:
MQ8: Yeah.
What would that temperature be? Warmer coolezi
E:
MQ8: It would be 5.
Would they feel the same?
E:
MQ8: No, the iron will be really cold.
What is going on?
E:
MQ8: The way that it can, the metal stove, the way it conducts the heat away from your body very rapidly, so it
feels really really cold.
So it is conducting the heat away from my body?
E:
MQ8: Yeah.
What if I touch the wood pile?
E:
MQ8: It would probably be cold but not as cold, but it wouldn't be so free7 g though.
So what is happening?
E:
MQ8: The wood is not a very good conductor, so it conducts the heat away from your body to various rates, so
like if you just touched it, it wouldn't feel that cold because it is not a very good conductor.
MQ8:
On the next question, which asked for the best cover for a warm object that would allow
someone to feel how warm it is, she chose Saran Wrap, but couldn't explain why. At first she
said it was a conductor then changed her mind. She chose aluminum foil to best feel the coolness
of cold objects and gave the explanation, "You could feel it with foil because it is a conductor. It's
a good conductor so when you feel it, it feels cold because it is conducting the heat away from
your body rapidly" Most students can explain heat flow in one direction long before they can
explain it in the other, but for most students it is more difficult to explain why objects feel cold.
She had the opposite problem as demonstrated in her response to who is right about the
tempetature of strips of metal and wood in the hot trunk of a car.
27
MQ8: Neither of us. They are both the same temperature because they've both been in the same surround.
Does that make sense?
E:
MQ8: Yeah, it is in the same surroundings so it can't be two different temperatures.
E:
Why not?
MQ8: Because they are both in the same surroundings, and they can be insulators and conductors, but they will be
the same temperature because... I don't know.

But they won't feel the same?
E:
MQ8: No, the wood will feel warm but not as warm as the metal.
How do I feel the sensation of heat?
E:
MQ8: If it is a good conductor and you toach it, then... I can always get the cold but not the hot... if it's hot,
maybe it absorbs the heat... I don't know.
Her inability to explain why objects feel different appears to affect her understanding of
thermal equilibrium. In Interview 3 after being asked about the temperature of objects in a
laboratory oven overnight, she stated:
MQ8:
They arc probably all the same temperature.
Why probably?
MQ8: Actually, they might be the same temperature because they are all the same surmundings...You know, they
are all in the same oven, they have all the same amount of heat energy added to them.
Do they feel the same?
E:
MQ8: No.
What might you fmd when you touch them?
E:
MQ8: The metal would probably feel the warmest, the glass would probably feel hot, the paper or the asbestos
stuff would be warm, and I don't know what that is...well, the glass is the same temperature as the beaker
E:
since it is still glass.

Are you real sure that they'd all be the same temperature?
MQ8: I am real sure.
Why is it if they are all the same temperature, they would feel different?
E:
MQ8: The metal is a conductor so when you touch it, it conducts heat away from your body...wait, no...[end of
tape)...I am just going to start over, they feel differently because they...actually maybe they couldn't be the
same temperature, because some things take more heat energy to heat them than cthers. So actually, I kind of
change what I said.
So they are not the same temperature in spite of the fact they they've been there over night?
E:
MQ8: Maybe, yes, it's possible.
E:
She is trying to put pieces of knowledge together and have them be consistent with her own
experiences. She has constructed an explanation for why metals feel cold but has been
unsuccessful is understanding heat flow from hot objects. Later in the same question she is
struggling to explain why the metal objects feel hotter. She finally decides they must have more
heat energy.
Let's imagine that this (points to metal spatula) has been there over night and I touch it. What happens to
make me say it is hot?
MQ8: The heat, it's flowing out of...well, it is conducting the heat and so...when you touch it, the, it'd feel hot.
The metal is conducting from where to where?
E:
MQ8: From itself...I don't know what to think...I don't know what to say, except that it's conducting heat maybe
from your body into the metal and then...I don't know...
So if I touch this and heat is conducted from my body into this, I am going to say that it is hot.
E:
E:
28
MQ8: Yes.
E:
If I touch the glass, it wouldn't feel as hot.(Editors note: she previously stated this.) So what is different
about the glass?
MQ8: It is probably not that good of a conductor but the metal is, so...
E:
So what happens that it doesn't feel quite as hot?
MQ8: I.t.ss heat energy.
E:
What?
MQ8: There is more heat energy in the metal...
E:
Does that make sense?
MQ8: Yes.
Progressing students encompass both students who may and may not observe
inconsistencies in their prediction or explanation responses. Their similarity comes in their
responses to either self-noted or interviewer-noted inconsistenciesresolution is difficult. They

may engage in sufficient reflection that knowledge reorganization occurs and new representations
develop, but this is a difficult and lengthy process. Additionally, insights are typically limited in
scope and students are often not able to apply them in situations where underlying principles are
similar but which differ in surface features. Progressing students knowledge reorganization may
not persist throughout a single interview or be carried from one interview to another. This is
demonstrated in the following excerpts.
CE4 relied on experiences to make prdictions throughout the interviews and had difficulty
seeing beyond the surface features of a question. He began the CLP cuniculum with typical
intuitive conceptions. He chose "tin foil" to keep the soda cold for lunch because, "Tin foil keeps
things the same temperature." In a later question on the metal and Styrofoam plates at room
temperature, he states, "Metal attracts heat." He later states, "wool has a lot of heat energy" and
"wool gives out heat energy." In Interview 1 he said the metal spoon would be at a lower
temperature than the wooden spoon in the 65C oven because metal was a good insulator and wood
was a good conductor. In a later question he was asked if all the objects in the room were the same
temperature. He responded yes and cited the Probing Your Surroundings experiment.
CE4: Well, before the experiment., I didn't really think about it but after it I knew they were the same.
E: Did the experiment make sense?
CM: Yeah, pretty much if they've been here for a long time.
E: When touch the things in room, don't feel same. Why don't feel the same if they are the same?
CM: Bo:cause of your uhm, body heat when you touch it because you're always warm, because you've only been
here a few minutes.
E: OK, so if touch this and I touch this, ...so what do you think?
CE4: Well, this (metal) would feel colder,but it's room temperature.
E: It really is room temperature?
CE4: Yeah, I think so.
E: Why do you think it feels colder?
CE4: Cause metal, well, I think cause the hot air rises and this stuff would kind of be colder, but it's all the same
temperature, I guess. I'm confused.
(skip ahead)
29
E: This question asks about a metal and Styrofoam plate in same room. You said the metal plate would be warmer
because, "Metal attracts heat."
CFA: (Reads) OK. I think that the metal plate should be wanner. Uhm, because, let me think this over a minute.
OK, the metal plate would absorb the heat and they Styrofoam, wait, I guess. The metal plaw would absorb the
heat around and the Styrofoam plate would just stay.
E: So the metal plate is warmer?
CFA: Yeah.
E: So if had metal plate and Styrofoam plate in here and had been sitting here since last night, would they be the
same or different temperatures?
CE4: Oh, overnight, I think they'd be the same temperatures.
E: But you were saying the metal plate would be warmer?
CM: I just thought like a couple of hours. I thought that they'd be in a freezer or something like that.
E: OK, so if they had been in freezer and I took the two plates out and put here, the metal plate would be warmer?
CM: Yeah, cause the Styrofoam would insulate it.
E: Insulate what?
CFA: The cold air that it just came out of.
E: Front the freezer?
CFA: Yeah.
E: OK and you said if they'd been here overnight,
CFA: They'd probably be the same. If they had enough time to get (inaudible).
E: OK. If I touched them, feel the same?
CE4: No, but if you got there temperature, they'd be the same.
(skip ahead)
E: This question was about shoveling snow.

CE4: rve shoveled snow before. rd probably want to use the wooden one.
E: Why?
CE4: Well cause the metal one would get, when I used one before it gets colder real fast and the wooden one seems
to be warmer.
E: Were _:te shovels different temperatures or the same temperature?
CE4: I think they're different when they're out in the snow. Cause when I've done it before the wooden one feels
warmer than the metal one.
E: If those two shovels were in a snowbank overnight, would their handles be the same or different temperatures.?
CM: I think they would be different because of what I've felt before.
E: Which one was warmer or coldell
CE4: I think the metal one would be colder and the wooden one would be warmer.
This student's inability to apply thermal equilibrium across problems is not unusual. Nor
is he usual in thinking of aluminum as an insulator. He later states that orange juice would stay
colder in an aluminum cup than it would it a Styrofoam cup because aluminum has more insulation
on it.
In Interview 2, CE4 makes predictions consistent wit.h the concept of thermal equilibrium,
and while he is unable to explain why objects feel differently, evidence of some knowledge
construction is present.
What kinds of things feel colder?
CFA: Metal does, usually, I think.
So if I touch metal, it feels colder, what's happening?
E:
CE4: I guess because it can absorb the heat energy and then. . .
The heat energy from what?
E:
CE4: From the room, I guess. I am confused. I can feel this and it feels pretty warm and this feels colder, but it is
actually the same.
E:
But you are not sure why?
CE4: Yes.
E:
30
Did you do the potato or Coke experiment? Did you remember doing the principle? Do you remember what
your principle was?
CEA: No, I don't. I think it was something to do with the temperature. I am not sure.
E:
What do you remember about the experiments themselves, when did you discover about the potato
experiments?
CE4: We found that the wool was beuer because it is a better insulator to keep the potato warm and that aluminum
foil, when we went on, it got all the heat energy from the potato until it got the the same temp.
E:
So what does that tell you? What do insulators and conductors do?
CE4: Well, an insulator will keep something cold or hot, as long as it possibly can, and then a conductor will take
the heat from it from another object, and then it will even out to then.
E:
(skip ahead)
E:
If we imagine wrapping something in an insulator, the insulator does what?
CE4: It keeps the thing cold or warm. It keeps the temperature.
In the Interview 3 he is still struggling with his knowledge of thermal equilibrium and his
intuitive conceptions about metals becoming warmer than their environment He is asked about the
temperature of objects inside a 150C laboratory oven overnight.
I think the metal spatulas would be the hottest.
CE4:
E: You think it'd be about how hot?
For overnight if you left it? Almost 200, maybe.
CE4:
E: So it'd be hotter than the oven?
No, I don't think so, actually. It'll probably only be about 160 or something like that.
CE4:
E: So it would be hotter than the oven?
CE4:
Yeah.
E: But not a lot hotter than the oven? Okay, but it'd be hotter than the oven. What about the other things in there?
I think all the glass stuff would be almost even with the oven.
CE4:
E: What about the asbestos pad?

CE4:
What's it made out of?
E: Asbestos. [explanation]
I think it'll be even, too, with the oven.
CE4:
E: But you think the metal would be hotter than the oven.
CFA:
Not too much.
E: But not that much hotter, maybe 10 or 15 degmes, you said.
Not that hot, actually, like 10.
CE4:
E: How do you think the metal gets hotter than the oven?
I think . . . okay . . .
CE4:
E: What are you thinking?
I guess I kinda messed up on the metal because I didn't think about how long they would be there.
CE4:
What if they were there two days, would that make a difference?
CE4:
Yeah, I think they'd all pretty much all be the same temperature.
E: What if they were there two hours?
CE4:
I would think the metal would be hotter.
E: Hotter than the other things?
CE4:
Yeah.
E: What about the oven?
CE4:
I don't think so, well, yeah, a little bit
E: You were originally saying that you though, the metal things were hotter than the oven and then it looked like
you had a conflict in your own mind when I asked you how it would get hotter than the oven. So what was
going on in your mind at that point?
CE4:
I was thinking about time, then I got confused.
E: So you think if the time is right that the metal can get hotter than the oven that it's in?
CE4:
Yeah.
E:
E: How would that happen?

Actually now that I think about it, I don't think it can.
CE4:
E: Why do you think it can't now?
31
Because inside when you close it, that would be the temperature right there. I don't think it could let any
CE4:
hotter than the oven is itself.
E: So why do you think it can't get hotter than the oven?

CE4:
(inaudible)
E: What is it that doesn't allow things to get hotter than the even?
Because I guess when you put in the metal and all that stuff, it's not even close to 150 degrees so I don't
CL4:
think it could heat up more than the oven itself.
E: Can things ever get hotter than the oven that they're in?
I don't know. I'm not sure about that.
CE4:
E: I don't mean things that genc-14 heat but just normal, inanimate objects.
CE4: Not like rue.
E: No, :401t rike rue, but the things you see in front of you. If you put any of them in an oven, could they get
1witter than the oven?
I cling% think so.
CEC
It looked like you were thinking one way and then you kind of changed the way you thought. What made you
change your mind?
After I thought about it for awhile . . .
CE4:
E:
After his refloction and reconsideration of thermal equilibrium, he overextends the concept
to include objects also feeling the same. It is almost as though he needs explicit help in focusing
on the important features of the problem. Another possibility is that he doesn't pay much attention
to what is being said or what he is saying. A third possibility is that this is very difficulty for him
and requires a considerable amount of work. His comments about class discussion may indicate
he has a hard time listening and understanding. His discussion of what he perceives as helpful
also provides insight. He finds the simulations most useful. These are quite concrete and related
to his daily life.
E: If you touched all these things, the minute you opened the oven door, what do you think you'd find then?
They'd all feel the same, I think.
CE4:
E: You think they'd feel the same, too?

Yeah.
CE4:
E: So if I touched the metal spatula or the asbestos pad, you think they'd feel exactly the same?
Oh wait, no. The metal one would feel hotter.
CE4:
E: Why?
Because they'd be giving out heat to your fingers when you touched it.
CE4:
E: Better than the asbestos?
Yeah.
CE4:
What
is it about metal that lets it do that?
E:
Just the material it's made out of.
CE4:
E: Is there a special name that you give to materials that do that sort of thing?
I don't know.
CE4:
E: Like what metals are called? Conductors?
Yeah.
CE4:
E: When you were thinking about things, you could tell that you sort of changed the way you were thinking about
it. Did you always think this way before you came into the class, or do you remember what you thought before
you came into the class?
Before this year I didn't really do anything with heat energy.
CE4:
E: So in a case like this, if sorneboAy asked you if all the objects in the oven were the same temperature, what
might you have said?
Probably said no, that scrne of them would have been hotter thrin others.
CE4:
E: Like you said at first? So what in this class worked to help you understand these ideas?
When we did the computer stuff on the Macs, that helped me a lot.
CE4:
E: The experiments and the simulations both, or . . .
32
I guess the simula ions.

CE4:
E: The simulations were better than the experiments?
CM: Yeah.
E: What about class discussion?

CM: No, because I don't really understand Mr. K. a lot of times.
E: You don't undezstand what he's saying?
CE4:
It's just hard for me to key in on what he's trying to talk about because a lot of people tell me stuff like
how the class, like my friends tell me what's going to go on and then he says something different, then I get
lost.
I wonder if your friends misunderstood what he was going to say.
CM: Yeah.
E: Are they in h different subject?
CM: No, ihey have him but a different time, period.
E: So they tell you he's going to say one thing and he says something really different? Or he just talks about
different subjects.
CM: Yeah, they might have different lessons.
E: That happens. How about the principles? Have they been helpful?
CM: Yeah, I guess so, because I use them in my thinking, I guess.
E:
E: How do you use them in your thinking?

CE4:
It's there and then it clicks in my mind, and that's how I use it.
E: Do the principles help you apply ideas to new problems, or not?
Sometimes.
CE4:
E: Sometimes but not all the time.
CE4:
Yeah.
E: Do the prototypes help. . ., or do they just
. . .
CE4:
No.
E: So probably the most useful thing to you is the simulations?
CM: Yeah.
E: Or do a bunch of thing work together?
CM: I would say the simulations are the best.
Later in the interview when placing materials on the continuum line, he cited experiments
and simulations as explanations. He found the question very hard because, "I don't know
(mumbles), because some of these things, I wouldn't even think about using." His references for
predictions and explanacions are usually concrete experiences. Generalizations are difficult for this
student. During Interview 4 he a gave glod explanation of why conductors in the pizza oven
would feel warmer than other materials. When the question was placed in the context of a freezer,
his separate ideas about the cooling and warming of insulators became evident.
What is it that a conductor does?
It gives off heat energy, it gives it to whatever is... give it to its surround.
CE4:
E: What if a conductor was colder than the surround, would it still give off heat energy?
CM: So if you have it from the freezer?
E: Yes.
I guess, I don't know... I am confused. It would... i think it will melt.., well my first thoughts that it
CE4:
would take in heat energy but that goes against what I really think, I am not sure.
E: What happens if you put a chunk of metal on the counter for a long time?
It will get warmer.
CE4:
E: So if you put a cold thing in a warm surround...
CM: I think the heat will overcome the...
E: I ow is it going to get warmer?
By taking in heat energy fran the surround.
CE4:
E: What did you think of that second, what did you think first?
E:
33
Well, I thought the metal was a conductor, then when you said you are putting on a cold, then I didn't
know, I got mixed up...
E: Can it still be a conductor?
Yes, I think so.
CE4:
E: How is it being a conductor if it is colder than it's surroundings?
Well, I guess, it will take in heat from the surround and putting it in some...
CE4:
E: Can conductors do that?
Yes.
CE4:
E: Would they do that easily?
No, I don't think so...
CE4:
E: So they may warm up with more difficulty than they can cool off?
Yes.
CE4:
CE4:
It should be noted that his confusion did not extend to insulators. In the next bit of dialog,
he gave good descriptions of the behavior of insulators in warm and cold envirOnments. Later he
restated his idea that conductors lose heat more easily than they gain it. This could be related to the
powerful intuitive conception about aluminum foil/metals being able to stay cold. It also displays
that his knowledge consists of pieces that have not been integrated. When he receives assistance
cueing his pieces of knowledge, his understanding seems to improve. But useful knowledge is not
cued on a reliable basis. Support for this is found in the next question where he constructs a
container to keep hot objects hot and cold objects cold. He chose Styrofoam for the middle and
lined the outside with aluminum. His explanation for his choice follows:
I remember my mom always wrap the cans with that.
CE4:
z: So the aluminum will help?
Yes, sort of like the protector.
CE4:
E: Does the aluminum work beuer than the Styrofoam?
I don't think so, I think Styrofoam is better.
CE4:
E: Is aluminum a metal?
Oh yeah, I don't know... I think so.

What if we put something hot inside?
Well then, the hot thing will lost its heat energy slowly but... the container would heat up as a whole and
CE4:
the object will also cool down a little bit.
E: Will the object cool down fast or slowly?
I think it would cool down slow.
CE4:
E: What about the trap air?
I think it gets warmer, it goes with the containe....
CE4:
E: So the air will get as warm as the temperature of the container?
I think it will cool off as much as the container thing gets warm.
CE4:
E: What happens to the heat energy that passed through the Styrofoam and hits the aluminum on the outside?
CM: I think since it's a metal, I think I messed up. I shouldn't have done that, it would go out.
E: Would it?
Yes, I am pretty s'Ire.
CE4:
E: Maybe it wouldn't make a difference... what would we use in its place?
Nothing comes to mind.
CE4:
E: Maybe it's okay to just use Styrofoam?
I think you should have another outside, but I just can think of anything.
CE4:
E: What if we put a cold object in there?
I think that the heat energy from the container will be absorbed by the cold thing, so the cold thing will
CE4:
slowly heat up and then the cold that it had will just like... I don't know... I think slowly the air inside of it
will get a little cooler but it would be melting down slowly.
E: What about heat energy from the outside?
I don't think it could get in... well, I don't know.. I think some of it could get in... well, I don't think so..
CE4:
CM:
E:
34
E: A little or a lot?
A little if any...
E: You think this is good then?
CE4:
I am not sure...
E: Can you think of a better way to do it?
[pause] I don't... I don't know.., buy one from the store.
CE4:
CEA:
His uncertainly was evident in Interview 5 when he discussed his posttest responses.
When asked to explain his heat energy and temperature distinction he responded, "OK, I knew, I
knew this yesterday. I think I have a hlork on it." Later when reading one of his correct
responses, he made a noise and said, I don't know. It seems wrong." Unfortunately, he could
not explain further. He then stated, "Wait. This. I'm not sure. Now I'm confused. But I'm sure
of that answer there. I just throw in the fffst thing I think of." This is probably not a bad
representation of his process. He has developed new, useful intuitive conceptions, but his
responses are cued by a variety of stimuli and somewhat varied. There is substantial evidence that
while he has improved in his understanding of individual pieces of concepts, his knowledge lacks
coherence and robustness.

Progressing students develop the ability to make accurate predictions reasonable rapidly,
but explanations for those predictions and events develop more slowly. Their explanations
typically refer directly to analogous experiences with limited numbers of principled explanations.
Eighteen of thirty-three (or 54.7 %) students fell into this category
Overall characteristics:
Consistency in predictions and explanations is less important for Progressing students
but more important than it is for Converging Students. While consistency is of some concern,
Progressing students have difficulty integrating concepts ideas in spite of a desire to do so.
The explanations of Progessing students vary in their attentiveness to experimental

results and constructed principles but do change over time as a result of their classroom
experiments/simulations and discussions.

Student predictions and explanations rely heavily upon previous everyday and classroom
experiences rather than principles. As a result, discussion are specific in nature and there is a
tendency to match surface features of problems rather than underlying principles. Figures 7 and 8
provide the Level of Explanations analyses for two Progressing students. A case study of
Progressing Student ME 10 can be obtained from the author upon request.
35
Figure 7:
Level of Explanations Over Sequence of Tests and Interviews for

Progressing Student CE4
Thermal Equilibrium
Heat & Temperature
Differentiation
Heat Flow
1
Explanations
3
PreT
Int #1
Int #3
Int #2
Int #4
PostT
#5
Mixed Predictions,
Idiosyncratic Explanations
Mixed Predictions,
Explanations
Sequence of Test and Interviews

Explanations
Target Conceptions
S.equence of tests
and Interviews
Pretests
Interview #1
Interview #2
Interview #3
Interview #4
Posttest
terview
Insulation/
Conduction
Thermal
Equilibrium
Heat & Temperature

Distinction
Heat Flow
4-
4
4
5
2
2
4
4
4
--/
4
4
,
37
44
36

Progressing Student ME10
-110-
-6--
Thermal Equilibrium
Heat & Temperature
Differentiation
Heat Flow
LffiLia_EXPlit1311lan
1
Explanations
Mixed Predictions,
kiosyncratic Explanations
Mixed Predictions,
Explandons
Target Conceptions
4
Int St
PteT
Int Kt
Int 02
Int #4
PostT
Int 15
Explanations
Level of Ex lanation
Sequence of 'rests
and Interviews
Pretests
Interview #1
Interview #2
terview #
terview
--Posttest
Interview #5
Thermal
Equilibrium
Heat & Temperature

Distinction
Heat Flow
2
4
.
Insulation/
Conduction
37
Development of Undgrstanding
Oscillating Students. The Oscillating student category is used for students who oscillate
between more and less predictive views of thermodynamics, who lacks integration of their ideas
and makes little overall progress toward target conceptions. At the end of the MP curriculum
these students had incorporated son. tiew facts and experienced some reorganization of
knowledge. However, their predictions and explanations were still based based largely on action
knowledge and intuitive conceptions. Some were able to incorporate some experimental outcomes
into their intuitive conceptions which limitedly enhanced the predictive power of those conceptions.
Unlike the students with a progressing view of thermodynamics, these students did not gain more
predicti.*1 ideas as time goes on but simply moved from one set of ideas to another. They did not
develop an understanding of the use and value of principles. As a result, they consider
thermodynamic events individually with no need for overall coherence. This resulted in the
application of different conceptions to questions with the same underlying principles.
This is demonstrated in excerpts from interviews of oscillating student CQ8. On the pretest
question about a wrap for a cold soda she wrote, "foil. Foil will keep it isolated from the heat."
During Interview 1 she was asked to say a little more about her answer.
CQ8: It'll keep it like, its not free, all the hot air can't get to it. It's surrounded by something, you know.
E: Would anything that surround something work as well as foil? If you put it in a zip-lock bag or something like
that.
CQ8: Well, I think that foil, well, yea, I guess it kinda would, but I think foil because it has that stuff, that frost
stuff, like you get on your mirror. Then foil kind of protects it and foil is tighter than a zip-lock bag.
She then goes on to explain why foil would work to keep things hot.
CQ8: I think it would because you put stuff in the oven, like if you want to bake lasagna then you have it in your
glass thing and it'll bake with foil on top of it.
E: What is the purpose of the foil in a case like that?
CQ8: To keep it isolated at whatever temperature you want, whatever temperature it is.
Later in the interview, after she discussed insulators, she decided that aluminum foil was an
insulator. Still later, when discussing whether objects in the same room were the same
temperature, she was asked for an example for her "Cannot predict" answer. Her reply was
telling, "Let me think of an example, then I can understand it better."
A few minutes later she announced, "Yeah, I think I've come to a conclusion about metal
and wood. I think that metals gets either hotter or colder than wood, like quicker." This is an
excellent observation. It may be surprising that she has never noticed this before, but clearly
reflecting on these questions has cause her to reorganized some of her thoughts.
38
In Interview 2 she stated that the metal objects would be the coldest in the 5C ski cabin
because they were conductors (note: she did the Probing Your Surroundings Experiment before
Interview 1). Her explanation for why they were colder was, "They can't...like uh, they well,
they keep, the heat energy goes out of them quicker than insulators. The heat energy just goes,
flows out quicker and it can't keep things warm." While this appears to be the beeinning of an
developing an understanding of conductors, she has a very different response to the question about
iron and aluminum.
CQ8: (The iron gets hotter because) it's a better insulator.
E: OK. So if it's a better insulator, what happens that it is able to get hotter.
CQ8: It, uh, it can keep the hot, it can keep the heat in.
It should be noted that most students assume iron must be a better conductor than
aluminum if it became hotter. In Interview 3 she explains that all the objects in the laboratory oven
would be the same temperature as the oven or hotter. When asked how they could get hotter than
the oven, she responded, "The heat energy comes out of something, like let's say it's a hot fridge,
then it would get into the objects." Later in the interview when asked the best wrap for a cold
candy bar she replied:
CQ8: I am thinking cf like the wood and I remember we did all this stuff. I think either wool or aluminum.
E: Wool or aluminum. OK. Would one work better than the other?
CQ8: I am trying to remember our experiment.
E: Let's think about each of them. Why do yon think wool would work?
CQ8: Because it's a good insulator and a good conductor. But a really good insulator almost like a good conductor,
you know? I was just learning that because--I think we learned that because an insulator keeps things warm and
keeps the outside air not to touch it, you know? Well, not to get near it. And it keeps the energy in. And so
wool would keep the cold energy in there..., you know.
After a few more interchanges, she was asked:
E: You said something about an insulator also being a conductor. Can you say a little more about that?
CQ8: It's not a conductor, it just kind of does the same things as a conductor.
E: What does a conductor do?
CQ8: It keeps things cool without letting the heat energy and the air come in.
E: Do conductors also keep things warm?
CQ8: No, because the conductor makes it so heat energy can't come it, and that keeps it for cold things, but an
insulator is, keeps the cold energy not to come in, just to keep the heat energy in.
After CQ8 gave an explanation of how aluminum would work to keep the candy bar cold,
she was asked if she had any idea which would work better.
CQ8: Um...no, I don't.
E: Would they work the same?
CQ8: No, I think one would work better.
E: But you are not really sure which?
CQ8: I can't really remember.
E: So you are inclir Id to say one, but you can't remember which one?
39
CQ8: Yeah.
E: Can you reason your way through it and see which one might be better?
CQ8: I think aluminum would because it was a bettu-jt was more of a conductor. Yeah, so like I would guess
aluminum. Late, on a test, I would guess aluminum.
Later on a question about the temperature of a cold room in the tropics, she stated that metal
would feel cold because it was a better conductor and "could keep in the cold better than the wood
can." In the next question on the continuum line, at fsst she placed metal as a good insulator, then
placed it in the middle of the continuum line giving a confused explanation, "I think that the metal
would be like...um right here you know? Oh no, not an insulator. I mean conductor. Well,
actually I den't because it's not a good insulatorit's not a good conductor because it does keep in
the cold and.-so I'd say a good insulator." When asked what a good insulator was she stated it
was "great at keeping things warm," and conductors "keep the cold inside of it."
In interview 5 she used the same definitions saying insulators keep hot things hot and
conductors keep cold things cold. She was integrating some of her classroom experiences into her
definitions and rued, "wool is a good insulator and a good conductor." She then explained that it
could keep hot food hot and cold food cold. Later in the interview she correctly predicted that the
wax on the metal block would melt first. The exchange reveals her conflict between what she
thinks should be thermal equilibrium and her own experiences.
CQ8: Metal will melt the quickest because metal always feels the hottest so the wax will melt the quickest
E: So the metal will feel hotter than the block of wood or would it get hotter?
CQ8: It would feel hotter, but it would not get hotter.
E: If it is the same temperature, wouldn't the wax melt at the same time?
CQ8: Yeah. Except, I still think that the wax on the metal will melt first.
It should be noted that on the post test she selected aluminum foil to wrap the cold soda for
school lunch. Her explanation was, "aluminum foil is a good conductor." Her posttest responses
represented a development of intuitive conceptions about insulators and conductors that she did not
have prior to taking the course. Additionally, she had incorporated experimental outcomes to the
extent that she knew that wool was effective at keep things hot and cold. Her explanation was not
useful in understanding future problems, but it did allow effective predictions about wool. Her
understanding of thermal equilibrium had progressed so that she made good predictions on all
thermal equilibrium posttest questions. While her explanations did not represent much in the way
of generalization, they were essentially correct. This is demonstrated by her explanation for why
the spoons in the 65C oven were both 65C: "I think in about that time they will become the
temperature of the oven." In spite of what seems like small progress, she effectively incorporated
the concept of thermal equilibrium and some additional action level knowledge about wool into her
existing knowledge. Her response to the request to, "describe something you learned in science
4i
40
classes which you could use to explain events outside of school" tells of her perceptions. "I could
use the stuff I learned about insulators and conductors a lot in my life."
Some oscillating students attempted to develop explanations for their action-level
knowledge and intuitive conceptions. However, the explanations were tied to their individual
interpretation of experiences in ways that did not allow the integration of new information. Their
idiosyncratic explanations were not altered by alternative classroom experiences. In fact, they
often made biased interpretations of experiments/simulations and misinterpreted patterns of
evidence to pre.serve intuitive conceptions or some constructed idiosyncratic explanation. This is
exemplified by the following selection of excerpts from student MQ7. On the pretest he named
aluminum as an example of a good conductor and wrote, "Aluminum can let electricity pass
through." When asked if this response was alright in Interview 1, he replied:

MQ6: Not really. I really didn't understand the question. (He mumbles something about Mr. K).... Aluminum. I
don't know about electricity. I don't know why I put that.
(editors note: later says aluminum a good conductor)
E: Why would you say aluminum is a good conductor?
MQ6: Cause it's uhm...uhm, it's just good. I don't know.
Students seemed to perceive these explanations as applicable in some cases and not in
others. As a result, they were unconcerned with responses that seem contradictory.
Contradictions were usually explained away with no apparent need for consistency. This may well
be based on some belief that science is not coherent (Songer, 1989). This is exemplified by a
response when the interviewer expressed general confusion with contradictory explanations to
similar questions. The student responded by patting the interviewer on the hand saying, "Don't
worry. I understand it."

Their reasoning was characterized by mapping of real world experiences onto questions
and responding to those questions based on those real world experiences. Oscillating students
made good predictions when such a match existed. Since their observations and explanations were
typically very concrete, it may be that seemingly straightforward thermodynamic explanations were
too abstract and unintelligible. As a result, principles had limited appeal and usefulness.
Oscillating students clung to alternative, simpler constructions which provided some explanation
for some events. As a result, these students appear to reject coherent explanations because they
cannot integrate their own intuitive conceptions with principles. Eight out of thirty-three students
(or 24.2%) fell into the Oscillating category.
41
Overall characteristics:
Consistency was local, not general for Oscillating students. Contradictions were never
noted by the students and were not acknowledged when noted by the interviewer. These
contradictions are instead seen as different problems requiring different solutions.
Oscillating students' reasoning was characterized by direct mappings of questions to

previous experiences. Explanations were derived from intuitive conceptions or intuitive
conceptions that evolved into idiosyncratic explanations that were perceived to apply to that
example.
Oscillating students' recall of experimental/simulated outcomes was often reconstructed
to supported their intuitive or idiosyncratic concepts. Figures 9 and 10 provide a Level of

Explanations analyses for two Oscillating students. A case study of Oscillating student MQ6 can
be obtained from the author on request.
42

Oscillating Student IVIQ6
Thermal Equilibrium
Heat & Temperature
Differentiation
Heat Row
Level of Exolanatlon
2
Mixed Predirlons,
lciosyncratic Explanations
Mixed Predictions,
Target Conceptions
Explanations
Int *1
Pn3T
Explanations
Int2
Int #3
Int #4
PostT
Int #5
s ano
--Sequence o ests
and Interviews
Thetests
Interview #1
terview #
Interview #3
terview
Thsttest
Conduction
uilibrium
Interview #5
Heat Flow
1
4
4
5
5
,
a eat : emperature
Distincfion
1
Explanations
....___4
2.
43
Oscillating Student CQ8
Thermal Equilibrium
Heat & Temperature
Differentiation
Heat Flow
1
Intuitive Concepticns
Explanations
Mixed Predictions,
Iciosyncratic Explanations
Mixed Predictions,
Target Conceptions
PieT
Int #1
Int #2
Int #3
PostT
Int S4
Explanations
Int #5
Explanations
equence 0 ests
and Interviews
Pretests
Interview #1
Interview #2
Interview ft3
Interview #4
' osttest
Interview #5
Heat F ow
Equilibrium
Heat . emperature
Distinction
2
4
2
2
4
Ins atio
Conduction
44
Comparison of Categories Using Posttest. Coding requirements for "correct" posttest

responses were not as rigorous as the requirements for achieving target conceptions in the Level of
Explanation Analysis. The primary source of this difference relates to what can be assessed from
an answer on a test compared to probing a student's response during an interview. For instance, a
student's response of "metals" as an example of a good conductor would be coded as correct. In
an interview, if that same response were probed and the discovery made that a student thought
metal was a conductor because it held cold, this response would not be considered correct. In spite
of these differences, a comparison of student categories and posttest means supported student
placement in those categories. Table 3, below, shows such a comparison. Since coding processes
and coding keys transcend semesters, this elves additional validity to the Levels of Explanation
method by which student understanding was assessed as well as to the categories into which
students were placed.
Table 3: ANOVA of Pretests and Posttest Means by Student Categories

Student Categories
Pretests Mean (Std.
Posttest Mean (Std.
Dev.)
Dev.)
98.7 (9.8)*
91.2 (6.9)**
77.2 (5.6)
Converging
!Dressing
18
61.4 (4.9)
61.6 (4.2)
Oscillating
60.9 (5.9)
* Converging Students' Posuest Means significantly better than Progressing and Oscillating students at p<0001
**Progressing students' Posttest Means significantly better than Oscillating students at p<.0001
The reliability of these student categories is further supported by comparisons of student
responses to the heat energy and temperature distinction question. The percentage of students
meeting strong criteria4 on this question has been used as a measure of student success in the CLP
curriculum for many semesters. Converging students performed better than Progressing students,
F (1, 31) = 3.9, p<.05. Converging students also performed better than Oscillating students, F (1,
31) = 8.4, p<.005. Additionally, Progressing students performed better than Oscillating students,
F (1, 31) = 2.1, p<.05.
4/Strong Criteria require a good differentiation between heat energy and temperature including one of the following
and an example: (a) intensive properties of heat energy and extensive properties of temperature; (b) a discussion of
heat flow which includes direction; or (c) a discussion of heat flow until thermal equilibrium is reached. The criteria
can also be met by two good examples or two good explanations using a, b, or c.
45
Intervention Effects
The vely act of interviewing students facilitated the development of understanding and
produced statistically significant differences in posttest scores. While the interviews were short
(appmximately 15-20 minutes per interview) and of limited number (4 interviews before the
posttest; one after), it is apparent that encouraging students to reflect deeply upon their responses
and explanations is beneficial in the development of understanding.
An effect size analysis shown in standard deviation units of test results over the course of
the 13-week curriculum demonstrates the improvement in student inderstanding (see Figure 11).
Students began with no significant differences in their pretest scores. Over time, interviewed
students progressed steadily in their test responses compared to the non-interviewed students. On
the posttest, significant differences were found in the areas of predictions and explanations of
thermal equilibrium, distinguishing heat energy and temperature, and heat flow.
Figure 11 shows that the interview produced no differences in r'ne area of insulation/
conduction probably because this already comprised a large part of both the curriculum and
students' classroom activities. Since the Spring 1989 semester, a major focus of the curriculum
has been insulation and conduction. Pilot studies showed that students' intuitive conceptions in the
the area of insulation and conducdon were persistent and presented substantial barriers to
understanding elementary thermodynamics (Lewis, 1987). This led to curricular reformulations

during the Spring 1989 semester: Simulations of everyday experiences related to insulation and
conduction were added (Lewis & Linn, 1989). Additionally, great emphasis was placed
throughout the curriculum on the concepts of insulation and conduction. As a result, all students
attained relatively high understanding in this area in subsequent semesters.
A consequence of the emphasis on insulation and conduction was that thermal equilibrium
received less attention and student performance in this area declined in both Spring and Fall of
1989. However, interviewed students again performed significantly better than non-interviewed
students in the area of thermal equilibrium. Further Figure 11 shows the progression in their
understanding from pretest along the way to posttest. These results are not surprising since each
interview contained at least one thermal equilibrium question about everyday situations. Students'
responses to these questions were then probed to determine their depth of understanding. As a
result, interviewed students spent more time considering thermal equilibrium and their own
understanding of that concept.
46
Figure 11: Effect Size by Comparing Interviewed Students with

Non-Interviewed Students
(All Students from Middle 50% on Pre-Tests)
Eavors
Concepts catered In interviews
Thermal Ecuilibrium
Si
-.12
.30
.51
.74
Ilaemy and Temperature
No Interview
Pre-test
Short Test 1
Short Test 2
Post-test
.46
Pre-test
Short Test 1
Short Test 2
Post-test
Heat Flow
.14
.55
Pre-test
Post-test
-.06
.12
.26
.19
-.26
.21
.41
Concepts not covered in Interviews
alittirditIBUL
Unequal Vohunes
.20
.25
Pre-test
Post-test
.21
Pre-test
Post-test
.31
Jnitial Temperature
Pre-test
Short Test 1
Short Test 2
Post-test
.44
Pre-test
Post-test
hancie in Temperature
.13
.16
Pre-test
Post-test
'Indicatias statistically significant at p
0.05
.21
Interview: Construct
Explanations or
I Answer Questions
47
It is also not surprising that no statistical differences arose for variables not covered in the
interviews, e.g. surfac e area, unequal volumes, initial temperatures, and temperature differences.
Most students entered the curriculum with good intuitive conceptions of these variables and their
understanding continued to increase over the course of the curriculum.
While using the CLP curriculum, students engage in the process of reflection. They give
explanations for predictions as well as experimental/ simulated outcomes. However, since these
predictions and explanations are "research group" constructions, individual students are engaged to
varying degrees. The process of interviewing students compels them to engage more deeply in the
process of reflection on an individual basis. Students participating in the interview process also
knew that their understanding of everyday thermodynamic phenomena would be probed regularly.
Students' anticipation of future interviews may have provided additional motivation to reflect
further on the questions asked during the interviews and to be pay geater attention to classroom
processes.
The interview also provided students the opportunity to compare their responses to related
questions within a given interview and check the consistency and coherence of these responses.
Since students often give contradictory responses to similar questions, the interview questions
were designed to illustrate inconsistencies in a student's knowledge. Students who perceived

those contradictions on their own would often engage in illuminating monologues where they
would sort out their thoughts (this will be discussed in detail in the section on cognitive growth).
As might be expected, these students were typically more successful at integrating their knowledge
and applying that knowledge successfull, to new problems. Other students who might normally
choose to ignore contradictions in their responses were presented with questions about those
contradictions. While they were free to say they had no explanation, the interviewer would return
to the contradiction in some new question in an attempt to encourage resolution. Alternately, some
students constructed ad hoc explanations or engaged in processes of reflection that were not
profitable but which were satisfactory to that student.
CONCLUSIONS AND IMPLICATIONS

Students participating in interviews were encouraged to engage in reflection by requiring
them to focus on their experimental/simulated outcomes as well as their own descriptions and
explanations of thermodynamic processes. In doing so, the interviews did not permit students to
give a single piece of information in response to a question. Instead, the interviews encouraged
and required students to continually reconsider their classroom and real-world experiences and
their own understanding of those experiences. In essence, the interview encouraged students to
;;
48
Development of Uulerstanding
reflect on their experiences and engage in sense-making processes. As a result of this reflection,
over time, most students' verbal representations of their ideas improved. This process, in turn,
appeared to encourage further knowledge reformulation.
Furthermore, the interviews, with their persistent requirements for explanations, promoted
and encouraged students to engage in more analytical processes. 'T" is, in turn, encouraged
students' generalization of their conceptions. The interviews, further, instilled in students a

propensity toward giving explanations. Over time, students increasingly incorporated explanations
into their predictions without being asked. During the interviews, it was not uncommon for a
student to say, "I knew you were going to ask me that!" in response to a "Why?" from the
interviewer. Later in the interviews, several students fmished giving a prediction and then
mimicked the interviewer by adding, "Why?" It was clear they expected to give explanations and
began to give them more spontaneously. This elaboration represents an additional step in
knowledge construction. Mindful, deliberate abstraction is a difficult process that takes time and
requires motivation by students. The interviews appear to have provided extrinsic motivation for
students to persist in reflective processes. This outcome is noteworthy because it shows that even
a short instructional intervention has the potential of encouraging students to engage in additional
reflection and elaboration of their ideas.
How Understanding ChangesThree Categories of Learners

All students proceeded to some extent by fits and starts, gaining some local reorganization
of knowledge, overextending new ideas, acquiring new pieces of knowledge, and integrating them
into their existing knowledge. These constructive processes proceeded at considerably different
rates, and revealed great diversity in students' understanding at the end of the CLP curriculum.
Some students engaged in a small scale reorganization of knowledge, while others showed
major restructuring of their concepts. Most students were somewhere between these extremes.
This middle group experienced major restructuring in some areas and was in the process of
knowledge acquisition and developing coherent understanding in other areas. Analysis of student
interviews suggested three general categories of students' development of understanding. These

categories are called Converging, Progressing, and Oscillating.
49
Student Categories of Knowledge Acquisition

The emergence of three categories in the development of students' understanding was
determined to some extent by differences in students' perceptions of the nature of the unit of
knowledge. Oscillating students perceived the unit to be at the level of actions or events, each
requiring different explanations. Thus phenomena, which were governed by the same principles,
but with differing surface features, were perceived as different domains (e.g. heating and cooling
would be considered different domains). As different domains, they required separate rules to
describe their behavior. Progressing students were beginning to combine these domains using
some underlying concepts and generalize beyond the surface features of phenomena. As an
example, they may have been able to apply the concept of thermal equilibrium to a variety of
circumstances, but were unable to explain why materials in those circumstances felt different
Their unit of knowledge was at the level of individual concepts (e.g. thermal equilibrium,
insulation, etc), but they had not integrated these concepts into a larger more cohesive view of
elementary thermodynamics. Converging students used all of the concepts in the curriculum as a
single domain that was expected to have internal coherence.
These very different views of thermodynamic problems yielaed very different approaches
to their solutions. Oscillating students tried to locate the appropriate piece of knowledge for a
given problem. Progressing students tried to locate the individual concept involved and determine
how it could be applied. Converging students examined the problem from a more global
perspective and were more able to see how individual processes contribute41, toward an given
outcome. Additional discussion of the characteristics of each of three categories of students

follow.
Converging Students.
inverging students are characterized by a robust and coherent
understanding of heat flow as a model for understanding elementary thermodynamics that begins to
converge with accepted scientific views. A key characterization of their reasoning processes was
their ability to resolve contradictions in their responses, reason through their in-and-out of-class
experiences, and decide which explanations were most consistent with other pieces of their
knowledge. Converging students cc itinued to add new information to their existing knowledge
and find that new knowledge was consistent with and reinforced the way they thought about
thermodynamic phenomena.
Converging students were more able to subject their intuitive conceptions to examinadon
and comparison with their experimental and simulated outcomes. In doing so, they constructed
new explanations for previous experiences. When Converging students used analoides in their
50
reasoning processes, they did not rely on surface similarities, but reasoned from underlying
thermodynamic processes. Their explanations for predictions in the interviews were more causal
and principled in nature. Some Converging students' behavior in the first interview suggested a
belief that all of their pieces of knowledge should fit together. Other Converging students
recognized this knowledge coherence later in the interview process. The emergence of this
awareness appears to accelerate students' progress toward knowledge integration. This hypothesis
is consistent with the findings of Songer (1989). She found that students' beliefs about the nature
of science were a predictor of the level of laiowledge integration they would attain.
Progressing Students. Students with a progressing understanding of thermodynamics are

incorporating new information to build a more robust and cohesive view of thermodynamics based
on the hez,i, flow model. These students are progressing towards the coherent understanding of the
students in the previous category. The progressing category, quite naturally encompasses great
diversity in student understanding. It includes students who have just begun to make sense of their
experiences and find general principles underlying a variety of phenomena as well as students who
are very close to developing the level of explanations required to fit into the category of converging
students.
The reasoning processes of Progressing students are best characterized by struggles to
resolve experimental/simulated outcomes and to develop relationships between those outcomes.
They advanced by fits and starts, often reached plateaus, followed unfruitful paths or, in some
cases, maintained two separate domains of knowledge (e.g. heating processes are different from
cooling processes). Overall they continued to progress toward a robust and coherent
understanding of thermodynamics.
An example of this knowledge construction is found in students' understanding of thermal
equilibrium. At first, students consider objects in different surroundings as different domains.

That is, different "rules" apply to objects in warni environments, cold environments, and at "room
temperature." The Probing Your Surrounding experiment is a powerful tool for enabling them to
simultaneously experience the identical temperatures and different feel of materials. This results in
the incorporation of facts relating to the temperature of objects at room temperature. Since this
matches students' intuitive belief that object.s in the same environment should be the same
temperature, it is more readily accepted and ultimately generalized to thermal equilibrium in
different surroundings. Their ability to explain the feel of materials comes later as they combine
their evolving facts and concepts of insulators/ conductors, heat flow, and thermal equilibrium.
Pieces of knowledge are constructed and integrated within the concepts of thermal equilibrium,
heat flow, and insulators/conductors. After integration within these concepts occurs, the concepts
51
must be linked before students can develop explanations for why conductors feel different than
insulators to the touch. This process is a difficult and lengthy one for Progressing students. The
level of integration that was just demonstrated gives insight into why many Propessing students
are never able to explain the feel of a variety of materials in different environments. Since
Progressing students' insights into problems are more limited in scope, they are often not able to
perceive situations whose underlying principles are similar, but differ in surface features.
Progressing students encompass both students who may and may not observe
inconsistencies in their prediction or explanation responses. Their similarity conies in their
responses to either self-noted or interviewer-noted inconsistenciesresolution is difficult.
Progressing students' knowledge reorganization may not persist throughout a single interview or
may not be carried from one interview to another.
Oscillating Student. The Oscillating student category is used for students who oscillate
between more and less predictive views of thermodynamics, who lack integration of their ideas and
make little overall progress toward target conceptions. At the end of the CLP curriculum, these
students had incorporated some new facts and experienced some reorganization of knowledge. At
the level of isolated knowledge, these students had made progress. As a result, their predictions
and explanations were still based based largely on action knowledge and intuitive conceptions.
Some were able to incorporate some experimental outcomes into their intuitive conceptions which
limitedly enhanced the predictive power of those conceptions. Unlike the students with a
progressing view of thermodynamics, these students did not gain more predictive ideas as time
went on, but simply moved from one set of ideas to another. They did not develop an
understanding of the use and value of principles. As a result, they considered thermodynamic
events individually with no need for overall coherence. Oscillating students were nnt concerned by
a lack of coherence or sense-making in their responses. This resulted in their application of
different conceptions to questions with the same underlying principles.
Implications of Student Categories. It is clear that learners develop understanding of

scientific principles in very different ways and at very different rates. The emergence of these
categories of student development of understanding in elementary thermodynamics has
implications for science instruction.
In general, curricula must have components that serve all categories of students. As an
example, the construction and use of principles was very difficult for Oscillating students. A
common response was, "They're too hardand they don't make sense." For these students to
make progress in understanding, experiments and simulations that relate to their real world
52
knowledge can be used to construct new intuitive conceptions. Since much of the reasoning
processes of Oscillating students is experience-based, experiments and simulations can be
encouraged as new referents. If explanations can be imbedded in these experiments and

simulations there is the possibility of increasing coherence. It is vital that a new way of thinking
about science must be made appealing if these students are to ultimately gain skill in detection and
application of principles. One way to expand their ideas about the nature of science is to model
problem-solving and sense-making processes. The development of simple models and prototypes
should appeal to these students. If this development is accompanied by student success in
predictions and explanations, Oscillating students may be led to see the usefulness and power of
generalizations.
These recommendations are equally valid for students who are beginning to use underlying
principles in their assessment of phenomena. While they will progress more rapidly in their
understanding, they still need representations to form bridges between their intuitive conceptions
that are the beginning of scientific principles and those principles. Furthermore, they need
experiences that require comparison of their intuitive conceptions that are not consistent with
scientific principles with experimental outcomes in real-world settings. Such experiences should
include experiments on and simulations of everyday phenomena so that they can begin the complex
and difficult process of knowledge integration. These students are able to construct and apply
principles with varying difficulty. Principle construction and integration activities should
accompany all experiments and simulations.
If instruction is to serve all students, it must have a variety of components that are
appropriate to the level of the learner, that encourage sense-making in exploring concepts and filet
facilitate knowledge integration. The absence of these components serves to reinforce many
students' ideas about the nature of science learning: "In science the best thing to do is just
memorize, because it isn't supposed to make sense." Students' fundamental views on the nature
of science must also be addressed in curricula. Included in this understanding of science must be
what constitutes evidence and the ability to compare evidence in order to reach conclusions.
This study demonstrated that the received view of learning does not describe the ways
students acquire knowledge. If the goal of instruction is to produce individuals with coherent and
robust knowle lge, telling students about science and science principles is not an effective method
for attaining these goals. This raises the question of what are appropriate and desirable levels of
stude
understanding. Is the goal to have students learn isolated facts that seem to have little
relevance in their lives? If so, current methods that provide students with information on a wide
breadth of subject matter as well as detailed coverage within those conceptsall in remarkably
53
short periods of time seems a reasonable way of attaining that goal. If the goal is for students to
develop an understanding of science and scientific principles, and be able to apply these concepts
to daily life, curient educational practices will not work.
Students' difficulty in knowledge acquisition and knowledge integration has been
demonstrated in this study. These difficulties existed in spite of a 13-week curriculum that
explored a few concepts in great depth. These difficulties existed in spite of a curriculum that
began with learners' intuitive conceptions and sought to assist them in knowledge construction
using real-time experiments and real-world simulations. These difficulties existed in spite of
efforts to engage students in deep reflection by having them engage in numerous integration
activities.
The implications for science instruction are obvious. Curricular emphasis must be placed
on depth rather than breadth and on understanding versus recall of facts. Curricula must provide
sufficient time to take into consideration both students' initial understanding of the subject matter
and their beliefs about the nature of science. Curricula must also provide sufficient time for a
variety of experiences that facilitate knowledge construction. Additionally, instruction that fosters
knowledge integration should be embedded in curricula. Two examples of this type of instruction
found in this study could be embedded in any curriculum: (1) encourage students to engage in deep
reflection and (2) instill in students a propensity to construct explanations. While these processes
require time, they should produce students whose knowledge is robust and coherent and who also
feel capable of understanding science and its relationship to their world.
Instruction that Fosters Knowledge Integration

Consideration of Intuitive Conceptions. The observation that students have persistent
intuitive conceptions is not new. What has been presented here are some insights into the natuie
and construction of these conceptions and causality for their persistence. Intuitive conceptions are
constructed by the very processes that are valuable to individuals and valued by sciencemaking
observations Ot'phenomena over a variety of circumstances and attempting to find generalizations
for those phenomena. Students' action knowledge and intuitive conceptions are the product of
many years of such observation and attempts, however limited, at sense-making of those
processes. This study has shown how resistant some intuitive conceptions are to change. This is
especially true when students do not understand and cannot apply alternative explanations.
Any instruction that wishes to facilitate the reorganization of students conceptions, in an
area where intuitive conceptions exist, would be wise to explicitly take those intuitions into
elt1
54
consideration. As this study demonstrated, students have great difficulty with ideas that do not fit
their world experience. Instruction must be connected to students' observations of the world and
have concerns for students' conceptions and their learning processes. It is, therefore, useful to
include real-world knowledge and experiences in instruction for several reasons. First, it
encourages the integration of knowledge instead of the isolated "school knowledge" and "real-
world" knowledge we find in students, nonscientists, and even a few scientists. Second, it
encourages students to develop alternative explanations for those intuitive conceptions that are
contrary to scientific principles. Instruction can ground useful notions of causality in students' real
world experiences. Additionally, it makes scientific knowledge easier to remember. Real world
experiences can serve as prototypes and cueing mechanisms for new intuitive conceptions as well
as more principled understanding.
Emphasis on Reflection and Explanations. When students engage in activities that

requires sustained reflection, they make greater cognitive gains and are more successful in
integrating knowledge. Additionally, instruction that instills a propensity for explanation
encourages students to engage in additional reflection that facilitates knowledge integration. The
process of sustained reflection is quite difficult for most students. This could be improved if
science curricula began to emphasize reflection as an integral part of the daily activities in the early
grades.
Students must engage in sense-making in science. Instruction must foster and encourage
student reflection on their own notions of science and the coincidence or conflict of those notions
with school science. Science curricula should emphasize students' construction of c. lanations of
phenomena on a regular basis. As deeper explanations become a part of a students' discourse in
science, they will engage in different learning pPocesses. This changes should motive changes in
the way students perceive phenomenamore analytically and with greater emphasis on underlying
processes.
Motivation for reflective process can be found by choosing activities for science insauction
that seems relevant to students' lives. A part of fostering reflection and construction of
explanations includes demonstrations of scientific processes and reasoning about those
demonstrations. Since demonstrations that relate to daily life are often most interesting and
remembered, one of the
als of science instruction should be to embed students' learning in real-
world coinexts. This not only creates functional learning environments but encourages students to
employ more scientific reasoning in their daily lives.
55
As shown in this study, mindful, deliberate abstraction of experiences into principles is
difficult for students. The process also takes time and requires student motivation. Curricula
contents must be appropriate to the cognitive demands for lmowledge integration. Extensive
coverage of fewer topics is more likely to provide the time required for students to develop
coherent and robust knowledge. Fewer topics are not enough. Science curricula must include
more than facts, experiments, and prototypes if students are to engage in the difficult process of
knowledge integration. Emphasis must be placed on activities that foster students' constniction
and application of more abstract and general scientific principles.
The question of how integration activities should be defined is worthy of future

consideration. Should integration activities been created from students in general or should they be
created bases more diagnostically for the, individual student or a combination of the two?
56
REFERENCES
Ausubel, D. P., Novak, J. D. & Hanesian, H. (1978). Educational psychology: A cognitive view.
(2nd. Ed.). New York: Holt, Reinhart, and Winston.
Baldwin, J. M. (1984). The development of the child and of the race. (Reprinted by Augustus M .
Kelley, 1968 Ed.). New York: MacMillan.
Caramazza, A., McCloskey, M. & Green, B. (1981). Misconceptions about trajectories of objects.
Cognition, 9, 117-123.
Carey, S. (1988). Conceptual differences between children and adults, Mind and Language, 3 (3),
167-181.
Carey, S. (1985). Conceptual change in childhood. Cambridge, MA: MIT Press.

Chaiklin, S. (1984). Reasoning with a physical-sciencticoncept. Unpublished dissertation,
University of Pittsburgh.
Chi, M. T. H. & Ceci, S. J. (1987). Content knowledge: Its representation and restructuring in
memory development. In H. W. Reese, & P. Lipsitt (Ed.), Advances in child development
and behavior (pp. 91-142). NY: Academic Press.
Clement, J. (1982). Students' preconceptions in elementary mechanics. American Journal of
Physics, 50, 66-71.

Clement, J., Brown, D. E. & Zietsman, A. (1989, March).Not all preconceptions are
misconceptions: Finding "anchoring conceptions" for grounding instruction on students'
intuitions. Paper presented at the American Educational Research Association, San
Francisco, CA.
Clough, E. E. & Driver, R. (1985). Secondary students' conception of the conduction of heat:
Bringing together scientific and personal views. Physics Education.
Dewey, J. (1902). The child and the curriculum. Chicago: The University of Chicago Press.
diSessa, A. A. (1988). Knowledge in pieces. Hillsdale, N. J.: Lawrence Erlbaum Associates.
57
diSessa, A. A. (1983). Phenomenology and the evolution of intuition. Hillsdale, N. J.: Lawrence
Erlbaum Associates.
Driver, R. (1981). Pupils' alternative frameworks in science. European Journal of Science
Education, 3 93-101.
Driver, R. and Bell, B. (1986). Student's thinking and learning of science: a constructivist view.
School Science Review, March, 443-456.

Erickson, G. L. (1979). Children's conceptions of beat and temperature. Science Education, 63
(2), 221-230.
Erickson, G. L. (1980). Children's viewpoints of heat: A second look. Science Education, 64 (3),
323-336.
Eylon, B. & Linn, M. C. (1988). Learning and instruction: An examination of four research
perspectives in science education. Review of Educational Research, 58 (3), 251-301.
Glaser, R. (1990). The reemergence of learning theory within instructional research. American
Psychologist, 45, 29-39.

Gunstone, R. F. (April, 1988).Some long-term effects of uninformed conceptual change. Paper
presented at the American Educational Research Association, New Orleans, LA.
Kirkpatrick, D. (1987). Technology in education: A middle school teacher's experience.
Technology and Learning, 1(3),

Law, N. W. V. (1988).Knowledge structures: Where can we find them? Paper presented at the
American Educational Research Association, New Orleans, LA.
Lewis, E. L. (1987). A study of knowledge elements present in the heat and temperature concepts
of adolescents and naive adults. University of California, Berkeley; Computer as Lab
Partner Project.
Lewis, E. L. & Linn, M. C. (1989, April). Heat energy and temperature concepts of adolescents,
naive adults, and experts: Implication.s for curricular improvements. Paper presented at the
Annual Meeting of National Association for Research in Science Teaching, San Francisco,
CA.
58
Lewis. E. L., Stern, J. L. & Linn, M. C. (1990, April). The effect of simulations on
thermodynamics understanding. Paper presented at the Annual Meeting of the American
Educational Research Association, Boston, MA.
Linn, M. C. (1983). Content, context, and process in reasoning during adolescence: Selecting a
model. Journal of Early Adolescence, 3(1-2), 63-82.
Linn, M.C.& Pulos, S. (1983). Male-female differences in predicting displaced volume: Strategy
usage, aptitude relationships and experience influences. Journal of Educational
Psychology, 75, 86-86.

Linn, M. C. (1987). Establishing a research base for science education: Challenges, trends and
recommendations. Journal of Research in Science Teaching, 24(5), 191-216.
Linn, M. C., Layman, J. & Nachmias, R. (1987). The cognitive consequences of microcomputerbased laboratories: Graphing skills development. Journal of Contemporary Educational
Psychology, 12 (3), 244-253.

Linn, M. C. (1988). Designing robust science instruction: A model for thermodynamics. Proposal
submitted to the National Science Foundation.
Linn, M. C., Songer, N. B., Lewis, E. L. & Stern, J. (1990). Using technology to teach
thermodynamics: Achieving Integrated Understanding. In D. L. Ferguson (Eds.),
Advanced Technologies in the Teaching of Math and Science. Berlin: Springer-Verlag.
Linn, M. C. & Songer, N. B. (in press). Teaching thermodynamics to middle school students:
What are appropriate cognitive demands? Journal of Research in Science Teaching,
McCloskey, M., Caramazza, A., & Green, B. (1980). Curvilinear motion in the absence of
external forces: Naive beliefs about the motion of objects. Science, 210, 1139-1141,
McCloskey, M. (1983). Naive theories of motion. In D Gentner & A. L. Stevens (Eds.) Mental
models. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Mokros, J. R. & Tinker, R. F. (1987). The impact of microcomputer-based labs on children's
ability to interpret graphs. Journal of Research in Science Teaching, 24(4), 369-383.
Osborne, R. J. & Wittrock, M. C. (1983). Learning science: A generative process. Science
Education, 67, 489-508.
59
Piaget, J. (1972). Science of education and the psychology of the child. New York: The Viking
Press.
Piaget, J. (1969, first published in 1929). The child' c conception of the world. Totowa, N. J.:
Littlefield., Adams & Co.
Piaget, J. (1952). The origins of intelligence in children. New York: National Universities Press.
Pines, A. L. and West, L. (1986). Conceptual understanding and science learning: An
interpretation of research within a sources-of knowledge framework. Science Education,
70, 583-604.
Posner, G. J., Strike, K. A., Hewson, P. W., & Gertzog, W. A. (1982). Accommodation of a
scientific conception: Toward,a theory of conceptual change. Science Education, 66 (2),
211-227.
Ranney, M. A. (1987). Changing naive conceptions of motion. Ph.D. Dissertation, University of
Pittsburgh.
Resnick, L. B. (1987). Learning in school and out. Educational Researcher, 16(9), 13-20.
Resnick, L. B. (1983). Mathematics and science learning: A new conception, Science, 220, 477478.
Schauble, L. (1988a). Pvcesses of constructive learning and intelference in children:
Understanding the causal structure of a microworld. Ph.D. Disserrack, T!:tachers College,
Columbia University.
Schauble, L. (1988b). Applying scientific thinking skills in self-directed exploration: A

microgenetic study of transfer of learning. Proceedings of the International Conference of
Intelligent Tutoring Systems, Montreal, Canada.
Schwab, J. (1971). The practical 3: Translation into curriculum. School Review, 81 (4), 501-22.
Shuell, T. J. (1988). The role of the student in learning from instruction. Contemporary
Educational Psychology, 13, 276-295.

Shuell, T. J. (1986). Cognitive conceptions of learning. Review of Educational Research, 56,
411-436.
t; 1
60
Shuell, T. J. (1982). Developing a viable link between scientific psychology and educational
practices. Instructional Science, 11, 155-167.

Songer, N. B. (1989). Promoting transfer in thermodynamics: Representing variables on a
continuum. Unpublished doctoral dissertation, University of California, Berkeley.
Strike, K. A. & Posner, G. J. (1985). A conceptual change view of learning and understanding.
In L. H. T. West, & A. L. Pines (Ed.), Cognitive structure and Conceptual Change. New
York: Academic Press.
Tiberghien, A. (1980). Modes and conditions of learning. An example: The learning of some
aspects of the concepts of heat. Leeds, England: University of Leeds Printing Service.
Tiberghien, A. (1982). The development of ideas with teaching. S. F.: Academic Press.
Tschirgi, J. E. (1980). Sensible reasoning: A hypothesis about hypotheses. Child Development,
51, 1-10.
West, L. and Pines, A. L. (Eds.). (1985). Cognitive structure and conceptual change. Orlando,
FL: Academic Press.
White, R. 1'. & Tisher, R. P. (1988). Research on natural sciences. In M. C. Wittrock (Ed.)
Handbook of research on teaching. New York: Macmillian.
Wittrock, M. C. (1974). Learning as a generative process. Educational Psychologist, 11, 87-95.
61

Understand Termodinamics

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Understand Termodinamics

Uploaded by

Copyright:

Available Formats

DOCUMENT RESUME

Lewis, Eileen Lob

This study investigates how students participating in

U.S. DEPARTMENT OF EDUCATION

'PERMISSION TO REPRODUCE THIS

EDUCATIONAL RESOURCES INFORMATIO4

a This document has been reproduced as

The ERIC Facility has assigned

In our iudgment, this document

Points or view Of opinions stated in this docu

SCOPE OF INTEREST NOTICE

TO THE EDUCATIONAL RESOUHCES

The Development of Understanding in Elementary Thermodynamics

Eileen Lob Lewis

The Nature of the Learner

knowledge is investigated. This knowledge revision occurs during a 13-week thermodynamics

Students' Existing Knowledge

amalgamation of actions and observations, they provide no I, ;1 of explanation beyond

understanding of thermodynamic phenomena.

Intniflye_fdactatim prototypic Circumstances auprising_Faknaigni

Metals absorb, hold, or

Touching a variety of objects,

Use of aluminum foil for

Metals absorb, hold, or

Touching a variety of objects,

Use of aluminum foil for

Conductors conduct heat

Insulators conduct heat

Conductors insulators don't feel hot

Touching a variety of objects

Wool warms things up Surround any object in wool

Wool cannot be used as an

Table 1. A list of intuitive -nceptions, their characteristics, prototypic circumstances,

Process of Knowledge Integration

This position is supported by research describing a number of important conditions that

Pretest, Part 1 Mean

Middle 50% of Students

Whole Population (N=180

Pretest, Part 2 Mean (Std

behavior problems. Additionally, any knowledge reorganization or reformulations promised to be

Figure 1: Sample Questions from Fall 1989 Semester Pretest/Posttest'

Give an example that explains your answer.

4a) Give an example of a good insulator.

b) What is the main reason for your answer.

b) Give the main reason for your answer.

a) Which shovel handle would you rather use? (circle one)

either shovel handle would he okay

b) Fill in the blanks to make a principle that applies to these shovels.

b) What is the main reason for your answer?

the peas would be warmer than the water.

As a means of studying the effect of various factors on the development of understanding,

Figure 2: SAMPLE INTERViEW QUESTIONS

(Posner, et al., 1982).

Level of Explanations Analysis. The Level of Explanations Analysis was developed to

Figure 3: Level of apPanations Used to Analyze

2Encoding New Facts without Explanationsstudents' responses display the

3Mixed Predictions, Idiosyncratic Explanationsstudents at this level give some

4Mixed Predictions, Explanationsstudents make both correct and incorrect

5Good Predictions, Mixed Explanationsstudents make excellent predictions and give

6Target Conceptionspredictions and explanations are consistent with target

flows from higher temperature

differs in amount needed

Figure 4: Diagram of Target Concepts in Elementary Thermodynamics